[PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Yinghai Lu]

	fix the 32bit with bigsmp
	fix 64 bit with irq migration
	ordering functions in io_apic_xx.c


to do:
    merge io_apic_xx.c

Thanks

Yinghai Lu

[PATCH 1/7] x86: some debug info for 32bit sparse_irq

[Yinghai Lu]

so could figure out who get interrupt, but vector_irq is not initialized.

Signed-off-by: Yinghai Lu  <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/io_apic_32.c |    8 ++++++--
 arch/x86/kernel/io_apic_64.c |    2 ++
 arch/x86/kernel/irq_32.c     |    4 ++--
 3 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c
index 9e111d6..429889d 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -1120,8 +1120,12 @@ next:
                         cfg->move_in_progress = 1;
                         cfg->old_domain = cfg->domain;
                 }
-                for_each_cpu_mask_nr(new_cpu, new_mask)
-                        per_cpu(vector_irq, new_cpu)[vector] = irq;
+		printk(KERN_DEBUG "assign_irq_vector: irq %d vector %#x cpu ", irq, vector);
+		for_each_cpu_mask_nr(new_cpu, new_mask) {
+			per_cpu(vector_irq, new_cpu)[vector] = irq;
+			printk(KERN_CONT " %d ", new_cpu);
+		}
+		printk(KERN_CONT "\n");
                 cfg->vector = vector;
                 cfg->domain = domain;
                 return 0;
diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c
index a11515f..4d1b02d 100644
--- a/arch/x86/kernel/io_apic_64.c
+++ b/arch/x86/kernel/io_apic_64.c
@@ -1392,6 +1392,8 @@ __apicdebuginit(void) print_IO_APIC(void)
 	printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid);
 	printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
 	printk(KERN_DEBUG ".......    : physical APIC id: %02X\n", reg_00.bits.ID);
+	printk(KERN_DEBUG ".......    : Delivery Type: %X\n", reg_00.bits.delivery_type);
+	printk(KERN_DEBUG ".......    : LTS          : %X\n", reg_00.bits.LTS);
 
 	printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)&reg_01);
 	printk(KERN_DEBUG ".......     : max redirection entries: %04X\n", reg_01.bits.entries);
diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c
index 63b4f6c..30b58b3 100644
--- a/arch/x86/kernel/irq_32.c
+++ b/arch/x86/kernel/irq_32.c
@@ -237,8 +237,8 @@ unsigned int do_IRQ(struct pt_regs *regs)
 
 	desc = irq_desc(irq);
 	if (unlikely(!desc)) {
-		printk(KERN_EMERG "%s: cannot handle IRQ %d vector %#x\n",
-					__func__, irq, vector);
+		printk(KERN_EMERG "%s: cannot handle IRQ %d vector %#x cpu %d\n",
+					__func__, irq, vector, smp_processor_id());
 		BUG();
 	}
 
-- 
1.5.4.5

[PATCH 2/7] x86: remove union about dest for log/phy

[Yinghai Lu]

let user decide the bits.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/io_apic_32.c |   18 +++++++-----------
 include/asm-x86/io_apic.h    |   16 ----------------
 2 files changed, 7 insertions(+), 27 deletions(-)

diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c
index 429889d..2d504fa 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -1230,7 +1230,7 @@ static int setup_ioapic_entry(int apic, int irq,
 
 	entry->delivery_mode = INT_DELIVERY_MODE;
 	entry->dest_mode = INT_DEST_MODE;
-	entry->dest.logical.logical_dest = destination;
+	entry->dest = destination;
 
 	entry->mask = 0;                                /* enable IRQ */
 	entry->trigger = trigger;
@@ -1342,7 +1342,7 @@ static void __init setup_timer_IRQ0_pin(unsigned int apic, unsigned int pin,
 	 */
 	entry.dest_mode = INT_DEST_MODE;
 	entry.mask = 1;					/* mask 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;
@@ -1431,19 +1431,15 @@ __apicdebuginit(void) 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;
 
 		entry = ioapic_read_entry(apic, i);
 
-		printk(KERN_DEBUG " %02x %03X %02X  ",
-			i,
-			entry.dest.logical.logical_dest,
-			entry.dest.physical.physical_dest
-		);
+		printk(KERN_DEBUG " %02x %02X  ", i, entry.dest);
 
 		printk("%1d    %1d    %1d   %1d   %1d    %1d    %1d    %02X\n",
 			entry.mask,
@@ -1723,7 +1719,7 @@ void disable_IO_APIC(void)
 		entry.dest_mode       = 0; /* Physical */
 		entry.delivery_mode   = dest_ExtINT; /* ExtInt */
 		entry.vector          = 0;
-		entry.dest.physical.physical_dest = read_apic_id();
+		entry.dest	      = read_apic_id();
 
 		/*
 		 * Add it to the IO-APIC irq-routing table:
@@ -2188,7 +2184,7 @@ static inline void __init 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;
diff --git a/include/asm-x86/io_apic.h b/include/asm-x86/io_apic.h
index 3b297e4..bdb64c3 100644
--- a/include/asm-x86/io_apic.h
+++ b/include/asm-x86/io_apic.h
@@ -87,24 +87,8 @@ struct IO_APIC_route_entry {
 		mask		:  1,	/* 0: enabled, 1: disabled */
 		__reserved_2	: 15;
 
-#ifdef CONFIG_X86_32
-	union {
-		struct {
-			__u32	__reserved_1	: 24,
-				physical_dest	:  4,
-				__reserved_2	:  4;
-		} physical;
-
-		struct {
-			__u32	__reserved_1	: 24,
-				logical_dest	:  8;
-		} logical;
-	} dest;
-#else
 	__u32	__reserved_3	: 24,
 		dest		:  8;
-#endif
-
 } __attribute__ ((packed));
 
 struct IR_IO_APIC_route_entry {
-- 
1.5.4.5

[PATCH 3/7] x86: make 32bit support per_cpu vector fix #1

[Yinghai Lu]

use cpu_online_map as target cpus for bigsmp. just like 64 bit doing

also remove some unused TARGET_CPUS macro.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 include/asm-x86/bigsmp/apic.h |   15 +++++----------
 include/asm-x86/es7000/apic.h |    3 +--
 include/asm-x86/numaq/apic.h  |    2 --
 include/asm-x86/summit/apic.h |    1 -
 4 files changed, 6 insertions(+), 15 deletions(-)

diff --git a/include/asm-x86/bigsmp/apic.h b/include/asm-x86/bigsmp/apic.h
index 0a9cd7c..1d9543b 100644
--- a/include/asm-x86/bigsmp/apic.h
+++ b/include/asm-x86/bigsmp/apic.h
@@ -9,22 +9,17 @@ static inline int apic_id_registered(void)
 	return (1);
 }
 
-/* Round robin the irqs amoung the online cpus */
 static inline cpumask_t target_cpus(void)
 {
-	static unsigned long cpu = NR_CPUS;
-	do {
-		if (cpu >= NR_CPUS)
-			cpu = first_cpu(cpu_online_map);
-		else
-			cpu = next_cpu(cpu, cpu_online_map);
-	} while (cpu >= NR_CPUS);
-	return cpumask_of_cpu(cpu);
+#ifdef CONFIG_SMP
+        return cpu_online_map;
+#else
+        return cpumask_of_cpu(0);
+#endif
 }
 
 #undef APIC_DEST_LOGICAL
 #define APIC_DEST_LOGICAL	0
-#define TARGET_CPUS		(target_cpus())
 #define APIC_DFR_VALUE		(APIC_DFR_FLAT)
 #define INT_DELIVERY_MODE	(dest_Fixed)
 #define INT_DEST_MODE		(0)    /* phys delivery to target proc */
diff --git a/include/asm-x86/es7000/apic.h b/include/asm-x86/es7000/apic.h
index bd2c44d..750afad 100644
--- a/include/asm-x86/es7000/apic.h
+++ b/include/asm-x86/es7000/apic.h
@@ -17,7 +17,6 @@ static inline cpumask_t target_cpus(void)
 	return cpumask_of_cpu(smp_processor_id());
 #endif
 }
-#define TARGET_CPUS	(target_cpus())
 
 #if defined CONFIG_ES7000_CLUSTERED_APIC
 #define APIC_DFR_VALUE		(APIC_DFR_CLUSTER)
@@ -81,7 +80,7 @@ static inline void setup_apic_routing(void)
 	int apic = per_cpu(x86_bios_cpu_apicid, smp_processor_id());
 	printk("Enabling APIC mode:  %s.  Using %d I/O APICs, target cpus %lx\n",
 		(apic_version[apic] == 0x14) ?
-		"Physical Cluster" : "Logical Cluster", nr_ioapics, cpus_addr(TARGET_CPUS)[0]);
+		"Physical Cluster" : "Logical Cluster", nr_ioapics, cpus_addr(target_cpus())[0]);
 }
 
 static inline int multi_timer_check(int apic, int irq)
diff --git a/include/asm-x86/numaq/apic.h b/include/asm-x86/numaq/apic.h
index a8344ba..0bf2a06 100644
--- a/include/asm-x86/numaq/apic.h
+++ b/include/asm-x86/numaq/apic.h
@@ -12,8 +12,6 @@ static inline cpumask_t target_cpus(void)
 	return CPU_MASK_ALL;
 }
 
-#define TARGET_CPUS (target_cpus())
-
 #define NO_BALANCE_IRQ (1)
 #define esr_disable (1)
 
diff --git a/include/asm-x86/summit/apic.h b/include/asm-x86/summit/apic.h
index c5b2e4b..0f68037 100644
--- a/include/asm-x86/summit/apic.h
+++ b/include/asm-x86/summit/apic.h
@@ -22,7 +22,6 @@ static inline cpumask_t target_cpus(void)
 	 */
 	return cpumask_of_cpu(0);
 }
-#define TARGET_CPUS	(target_cpus())
 
 #define INT_DELIVERY_MODE (dest_LowestPrio)
 #define INT_DEST_MODE 1     /* logical delivery broadcast to all procs */
-- 
1.5.4.5

[PATCH 4/7] x86_64: rename irq_desc/irq_desc_with_new - fix

[Yinghai Lu]

need to check if desc is null in smp_irq_move_cleanup

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/io_apic_64.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c
index 4d1b02d..e4f03cc 100644
--- a/arch/x86/kernel/io_apic_64.c
+++ b/arch/x86/kernel/io_apic_64.c
@@ -1950,6 +1950,9 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void)
 		irq = __get_cpu_var(vector_irq)[vector];
 
 		desc = irq_desc(irq);
+		if (!desc)
+			continue;
+
 		cfg = irq_cfg(irq);
 		spin_lock(&desc->lock);
 		if (!cfg->move_cleanup_count)
-- 
1.5.4.5

[PATCH 5/7] x86: make 32bit support per_cpu vector fix #2

[Yinghai Lu]

need to check if desc is null in smp_irq_move_cleanup

also migration need to reset vector too, so copy __target_IO_APIC_irq from 64bit

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/io_apic_32.c |  188 +++++++++++++++++++++++++----------------
 1 files changed, 115 insertions(+), 73 deletions(-)

diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c
index 2d504fa..e0f85b5 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -268,6 +268,7 @@ static struct irq_cfg *irq_cfg_with_new(unsigned int irq)
 	return cfg;
 }
 
+static int assign_irq_vector(int irq, cpumask_t mask);
 /*
  * Rough estimation of how many shared IRQs there are, can
  * be changed anytime.
@@ -437,6 +438,65 @@ static void ioapic_mask_entry(int apic, int pin)
 	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_cfg *cfg;
+	struct irq_pin_list *entry;
+
+	cfg = irq_cfg(irq);
+	entry = cfg->irq_2_pin;
+	for (;;) {
+		unsigned int reg;
+
+		if (!entry)
+			break;
+
+		apic = entry->apic;
+		pin = entry->pin;
+		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 void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	unsigned long flags;
+	unsigned int dest;
+	cpumask_t tmp;
+
+	cfg = irq_cfg(irq);
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cpus_and(tmp, cfg->domain, mask);
+
+	dest = cpu_mask_to_apicid(tmp);
+	/*
+	 * Only the high 8 bits are valid.
+	 */
+	dest = SET_APIC_LOGICAL_ID(dest);
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	__target_IO_APIC_irq(irq, dest, cfg->vector);
+	desc->affinity = mask;
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
+#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
@@ -591,44 +651,6 @@ static void clear_IO_APIC(void)
 			clear_IO_APIC_pin(apic, pin);
 }
 
-#ifdef CONFIG_SMP
-static void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t cpumask)
-{
-	struct irq_cfg *cfg;
-	unsigned long flags;
-	int pin;
-	struct irq_pin_list *entry;
-	unsigned int apicid_value;
-	cpumask_t tmp;
-
-	cfg = irq_cfg(irq);
-	entry = cfg->irq_2_pin;
-
-	cpus_and(tmp, cpumask, cpu_online_map);
-	if (cpus_empty(tmp))
-		tmp = TARGET_CPUS;
-
-	cpus_and(cpumask, tmp, CPU_MASK_ALL);
-
-	apicid_value = cpu_mask_to_apicid(cpumask);
-	/* Prepare to do the io_apic_write */
-	apicid_value = apicid_value << 24;
-	spin_lock_irqsave(&ioapic_lock, flags);
-	for (;;) {
-		if (!entry)
-			break;
-		pin = entry->pin;
-		io_apic_write(entry->apic, 0x10 + 1 + pin*2, apicid_value);
-		if (!entry->next)
-			break;
-		entry = entry->next;
-	}
-	desc->affinity = cpumask;
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-#endif /* CONFIG_SMP */
-
 #ifndef CONFIG_SMP
 void send_IPI_self(int vector)
 {
@@ -793,34 +815,6 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin)
 }
 EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
 
-/*
- * This function currently is only a helper for the i386 smp boot process where
- * we need to reprogram the ioredtbls to cater for the cpus which have come online
- * so mask in all cases should simply be TARGET_CPUS
- */
-#ifdef CONFIG_SMP
-void __init setup_ioapic_dest(void)
-{
-	int pin, ioapic, irq, irq_entry;
-	struct irq_desc *desc;
-
-	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);
-			desc = irq_desc(irq);
-			set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
-		}
-
-	}
-}
-#endif
-
 #if defined(CONFIG_EISA) || defined(CONFIG_MCA)
 /*
  * EISA Edge/Level control register, ELCR
@@ -2003,6 +1997,9 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void)
 		irq = __get_cpu_var(vector_irq)[vector];
 
 		desc = irq_desc(irq);
+		if (!desc)
+			continue;
+
 		cfg = irq_cfg(irq);
 		spin_lock(&desc->lock);
 		if (!cfg->move_cleanup_count)
@@ -2695,15 +2692,15 @@ void arch_teardown_msi_irq(unsigned int irq)
 
 #ifdef CONFIG_SMP
 
-static void target_ht_irq(unsigned int irq, unsigned int dest)
+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_DEST_ID_MASK);
+	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_DEST_ID(dest);
+	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);
@@ -2711,18 +2708,22 @@ static void target_ht_irq(unsigned int irq, unsigned int dest)
 
 static void set_ht_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
 {
+	struct irq_cfg *cfg;
 	unsigned int dest;
 	cpumask_t tmp;
 
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
-		tmp = TARGET_CPUS;
+		return;
 
-	cpus_and(mask, tmp, CPU_MASK_ALL);
+	if (assign_irq_vector(irq, mask))
+		return;
 
-	dest = cpu_mask_to_apicid(mask);
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
 
-	target_ht_irq(irq, dest);
+	target_ht_irq(irq, dest, cfg->vector);
 	desc->affinity = mask;
 }
 #endif
@@ -2925,6 +2926,47 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
 
 #endif /* CONFIG_ACPI */
 
+/*
+ * This function currently is only a helper for the i386 smp boot process where
+ * we need to reprogram the ioredtbls to cater for the cpus which have come online
+ * so mask in all cases should simply be TARGET_CPUS
+ */
+#ifdef CONFIG_SMP
+void __init setup_ioapic_dest(void)
+{
+	int pin, ioapic, irq, irq_entry;
+	struct irq_cfg *cfg;
+	struct irq_desc *desc;
+
+	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.
+			 */
+			cfg = irq_cfg(irq);
+			if (!cfg->vector)
+				setup_IO_APIC_irq(ioapic, pin, irq,
+						  irq_trigger(irq_entry),
+						  irq_polarity(irq_entry));
+			else {
+				desc = irq_desc(irq);
+				set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
+			}
+		}
+
+	}
+}
+#endif
+
 static int __init parse_disable_timer_pin_1(char *arg)
 {
 	disable_timer_pin_1 = 1;
-- 
1.5.4.5

[PATCH 6/7] x86: ordering functions in io_apic_32.c

[Yinghai Lu]

try to make functions in the same order to io_apic_64.c

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/io_apic_32.c |  176 ++++++++++++++++++++++--------------------
 1 files changed, 93 insertions(+), 83 deletions(-)

diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c
index e0f85b5..5639404 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -1029,23 +1029,6 @@ static int pin_2_irq(int idx, int apic, int pin)
 	return irq;
 }
 
-static inline int IO_APIC_irq_trigger(int irq)
-{
-	int apic, idx, pin;
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-			idx = find_irq_entry(apic, pin, mp_INT);
-			if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
-				return irq_trigger(idx);
-		}
-	}
-	/*
-	 * nonexistent IRQs are edge default
-	 */
-	return 0;
-}
-
 void lock_vector_lock(void)
 {
 	/* Used to the online set of cpus does not change
@@ -1190,6 +1173,23 @@ static struct irq_chip ioapic_chip;
 #define IOAPIC_EDGE	0
 #define IOAPIC_LEVEL	1
 
+static inline int IO_APIC_irq_trigger(int irq)
+{
+	int apic, idx, pin;
+
+	for (apic = 0; apic < nr_ioapics; apic++) {
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+			idx = find_irq_entry(apic, pin, mp_INT);
+			if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
+				return irq_trigger(idx);
+		}
+	}
+	/*
+	 * nonexistent IRQs are edge default
+	 */
+	return 0;
+}
+
 static void ioapic_register_intr(int irq, unsigned long trigger)
 {
 	struct irq_desc *desc;
@@ -1926,55 +1926,6 @@ static unsigned int startup_ioapic_irq(unsigned int irq)
 	return was_pending;
 }
 
-static void irq_complete_move(unsigned int irq);
-static void ack_ioapic_irq(unsigned int irq, struct irq_desc *desc)
-{
-	irq_complete_move(irq);
-	move_native_irq(irq, desc);
-	ack_APIC_irq();
-}
-
-static void ack_ioapic_quirk_irq(unsigned int irq, struct irq_desc *desc)
-{
-	unsigned long v;
-	int i;
-
-	irq_complete_move(irq);
-	move_native_irq(irq, desc);
-/*
- * 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
- */
-	i = irq_cfg(irq)->vector;
-
-	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
-
-	ack_APIC_irq();
-
-	if (!(v & (1 << (i & 0x1f)))) {
-		atomic_inc(&irq_mis_count);
-		spin_lock(&ioapic_lock);
-		__mask_and_edge_IO_APIC_irq(irq);
-		__unmask_and_level_IO_APIC_irq(irq);
-		spin_unlock(&ioapic_lock);
-	}
-}
-
 static int ioapic_retrigger_irq(unsigned int irq)
 {
 	send_IPI_self(irq_cfg(irq)->vector);
@@ -2040,13 +1991,61 @@ static void irq_complete_move(unsigned int irq)
 static inline void irq_complete_move(unsigned int irq) {}
 #endif
 
+static void ack_apic_edge(unsigned int irq, struct irq_desc *desc)
+{
+	irq_complete_move(irq);
+	move_native_irq(irq, desc);
+	ack_APIC_irq();
+}
+
+static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
+{
+	unsigned long v;
+	int i;
+
+	irq_complete_move(irq);
+	move_native_irq(irq, desc);
+/*
+ * 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
+ */
+	i = irq_cfg(irq)->vector;
+
+	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
+
+	ack_APIC_irq();
+
+	if (!(v & (1 << (i & 0x1f)))) {
+		atomic_inc(&irq_mis_count);
+		spin_lock(&ioapic_lock);
+		__mask_and_edge_IO_APIC_irq(irq);
+		__unmask_and_level_IO_APIC_irq(irq);
+		spin_unlock(&ioapic_lock);
+	}
+}
+
 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_ioapic_irq,
-	.eoi 		= ack_ioapic_quirk_irq,
+	.ack 		= ack_apic_edge,
+	.eoi 		= ack_apic_level,
 #ifdef CONFIG_SMP
 	.set_affinity 	= set_ioapic_affinity_irq,
 #endif
@@ -2094,11 +2093,6 @@ static inline void init_IO_APIC_traps(void)
  * The local APIC irq-chip implementation:
  */
 
-static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
-{
-	ack_APIC_irq();
-}
-
 static void mask_lapic_irq(unsigned int irq)
 {
 	unsigned long v;
@@ -2115,6 +2109,11 @@ static void unmask_lapic_irq(unsigned int irq)
 	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
 }
 
+static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
+{
+	ack_APIC_irq();
+}
+
 static struct irq_chip lapic_chip __read_mostly = {
 	.name		= "local-APIC",
 	.mask		= mask_lapic_irq,
@@ -2631,13 +2630,31 @@ static struct irq_chip msi_chip = {
 	.name		= "PCI-MSI",
 	.unmask		= unmask_msi_irq,
 	.mask		= mask_msi_irq,
-	.ack		= ack_ioapic_irq,
+	.ack		= ack_apic_edge,
 #ifdef CONFIG_SMP
 	.set_affinity	= set_msi_irq_affinity,
 #endif
 	.retrigger	= ioapic_retrigger_irq,
 };
 
+
+static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc, int irq)
+{
+	int ret;
+	struct msi_msg msg;
+
+	ret = msi_compose_msg(dev, irq, &msg);
+	if (ret < 0)
+		return ret;
+
+	set_irq_msi(irq, desc);
+	write_msi_msg(irq, &msg);
+
+	set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
+
+	return 0;
+}
+
 static unsigned int build_irq_for_pci_dev(struct pci_dev *dev)
 {
 	unsigned int irq;
@@ -2663,17 +2680,10 @@ int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
 	if (irq == 0)
 		return -1;
 
-	ret = msi_compose_msg(dev, irq, &msg);
 	if (ret < 0) {
 		destroy_irq(irq);
 		return ret;
-	}
-
-	set_irq_msi(irq, desc);
-	write_msi_msg(irq, &msg);
-
-	set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq,
-				      "edge");
+        }
 
 	return 0;
 }
@@ -2732,7 +2742,7 @@ static struct irq_chip ht_irq_chip = {
 	.name		= "PCI-HT",
 	.mask		= mask_ht_irq,
 	.unmask		= unmask_ht_irq,
-	.ack		= ack_ioapic_irq,
+	.ack		= ack_apic_edge,
 #ifdef CONFIG_SMP
 	.set_affinity	= set_ht_irq_affinity,
 #endif
-- 
1.5.4.5

[PATCH 7/7] x86: ordering functions in io_apic_64.c

[Yinghai Lu]

try to make functions in the same order

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/kernel/io_apic_64.c |   67 ++++++++++++++++++++---------------------
 1 files changed, 33 insertions(+), 34 deletions(-)

diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c
index e4f03cc..a14c29e 100644
--- a/arch/x86/kernel/io_apic_64.c
+++ b/arch/x86/kernel/io_apic_64.c
@@ -409,40 +409,6 @@ static bool io_apic_level_ack_pending(unsigned int irq)
 	return false;
 }
 
-/*
- * Synchronize the IO-APIC and the CPU by doing
- * a dummy read from the IO-APIC
- */
-static inline void io_apic_sync(unsigned int apic)
-{
-	struct io_apic __iomem *io_apic = io_apic_base(apic);
-	readl(&io_apic->data);
-}
-
-#define __DO_ACTION(R, ACTION, FINAL)					\
-									\
-{									\
-	int pin;							\
-	struct irq_cfg *cfg;						\
-	struct irq_pin_list *entry;					\
-									\
-	cfg = irq_cfg(irq);						\
-	entry = cfg->irq_2_pin;						\
-	for (;;) {							\
-		unsigned int reg;					\
-		if (!entry)						\
-			break;						\
-		pin = entry->pin;					\
-		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
-		reg ACTION;						\
-		io_apic_modify(entry->apic, reg);			\
-		FINAL;							\
-		if (!entry->next)					\
-			break;						\
-		entry = entry->next;					\
-	}								\
-}
-
 union entry_union {
 	struct { u32 w1, w2; };
 	struct IO_APIC_route_entry entry;
@@ -625,6 +591,39 @@ static void __init replace_pin_at_irq(unsigned int irq,
 		add_pin_to_irq(irq, newapic, newpin);
 }
 
+/*
+ * Synchronize the IO-APIC and the CPU by doing
+ * a dummy read from the IO-APIC
+ */
+static inline void io_apic_sync(unsigned int apic)
+{
+	struct io_apic __iomem *io_apic = io_apic_base(apic);
+	readl(&io_apic->data);
+}
+
+#define __DO_ACTION(R, ACTION, FINAL)					\
+									\
+{									\
+	int pin;							\
+	struct irq_cfg *cfg;						\
+	struct irq_pin_list *entry;					\
+									\
+	cfg = irq_cfg(irq);						\
+	entry = cfg->irq_2_pin;						\
+	for (;;) {							\
+		unsigned int reg;					\
+		if (!entry)						\
+			break;						\
+		pin = entry->pin;					\
+		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
+		reg ACTION;						\
+		io_apic_modify(entry->apic, reg);			\
+		FINAL;							\
+		if (!entry->next)					\
+			break;						\
+		entry = entry->next;					\
+	}								\
+}
 
 #define DO_ACTION(name,R,ACTION, FINAL)					\
 									\
-- 
1.5.4.5

Re: [PATCH 3/7] x86: make 32bit support per_cpu vector fix #1

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> +#ifdef CONFIG_SMP
> +        return cpu_online_map;
> +#else
> +        return cpumask_of_cpu(0);
> +#endif

unrelated sidenote - shouldnt the scheduler define cpu_online_map to 
cpumask_of_cpu(0) on UP, to avoid such ugly #ifdefs?

	Ingo

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> 	fix the 32bit with bigsmp
> 	fix 64 bit with irq migration
> 	ordering functions in io_apic_xx.c
> 
> 
> to do:
>     merge io_apic_xx.c

applied to tip/irq/sparseirq - thanks Yinghai, great work!

What do you think about Eric's abstraction suggestion, that irq_desc's 
structure should be hidden from all but kernel/irq/* code. I think 
that's a sane suggestion. We do it in mm/slab.c too: no external code 
knows the structure of 'struct kmem_cache'.

	Ingo

Re: [PATCH 3/7] x86: make 32bit support per_cpu vector fix #1

[Yinghai Lu]

On Fri, Aug 15, 2008 at 1:21 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>
>> +#ifdef CONFIG_SMP
>> +        return cpu_online_map;
>> +#else
>> +        return cpumask_of_cpu(0);
>> +#endif
>
> unrelated sidenote - shouldnt the scheduler define cpu_online_map to
> cpumask_of_cpu(0) on UP, to avoid such ugly #ifdefs?

copy that from mach_default...

should double check if cpu_online_map is cpu_mask_of_cpu(0) if !CONFIG_SMP

YH

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Yinghai Lu]

On Fri, Aug 15, 2008 at 1:27 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>
>>       fix the 32bit with bigsmp
>>       fix 64 bit with irq migration
>>       ordering functions in io_apic_xx.c
>>
>>
>> to do:
>>     merge io_apic_xx.c
>
> applied to tip/irq/sparseirq - thanks Yinghai, great work!
>
> What do you think about Eric's abstraction suggestion, that irq_desc's
> structure should be hidden from all but kernel/irq/* code. I think
> that's a sane suggestion. We do it in mm/slab.c too: no external code
> knows the structure of 'struct kmem_cache'.

current arch code use irq_desc directly..., if need to irq_desc to
them, need to provide set_irq(..., ...)

YH

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> On Fri, Aug 15, 2008 at 1:27 AM, Ingo Molnar <mingo@elte.hu> wrote:
> >
> > * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
> >
> >>       fix the 32bit with bigsmp
> >>       fix 64 bit with irq migration
> >>       ordering functions in io_apic_xx.c
> >>
> >>
> >> to do:
> >>     merge io_apic_xx.c
> >
> > applied to tip/irq/sparseirq - thanks Yinghai, great work!
> >
> > What do you think about Eric's abstraction suggestion, that irq_desc's
> > structure should be hidden from all but kernel/irq/* code. I think
> > that's a sane suggestion. We do it in mm/slab.c too: no external code
> > knows the structure of 'struct kmem_cache'.
> 
> current arch code use irq_desc directly..., if need to irq_desc to 
> them, need to provide set_irq(..., ...)

that's ok - we can do include/linux/irq_desc.h that isnt generally 
included by drivers via interrupt.h. Then later on we can eliminate that 
include file altogether.

here's the current field usage histogram:

      3 desc->action
      3 desc->depth
      9 desc->affinity
     15 desc->chip
     22 desc->status

desc->status: mostly IRQ_MOVE_PENDING related. irq-balancing code should 
be abstracted into kernel/irq/* perhaps?

desc->action / desc->depth / desc->action: i dont think those should be 
accessed by genirq arch irq code.

desc->chip: there's a get_irq_chip() method already. irq_chip is 
something that obviously is known to architecture code - it's the irq 
controller 'driver' that the architecture code provides.

	Ingo

Re: [PATCH 3/7] x86: make 32bit support per_cpu vector fix #1

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> On Fri, Aug 15, 2008 at 1:21 AM, Ingo Molnar <mingo@elte.hu> wrote:
> >
> > * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
> >
> >> +#ifdef CONFIG_SMP
> >> +        return cpu_online_map;
> >> +#else
> >> +        return cpumask_of_cpu(0);
> >> +#endif
> >
> > unrelated sidenote - shouldnt the scheduler define cpu_online_map to
> > cpumask_of_cpu(0) on UP, to avoid such ugly #ifdefs?
> 
> copy that from mach_default...

sure, that's the right approach - preserve existing code to reduce risk.

> should double check if cpu_online_map is cpu_mask_of_cpu(0) if !CONFIG_SMP

yeah. Separate clean-up patch i think.

	Ingo

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Ingo Molnar]

btw., one new thing i got in testing due to irq/sparseirq is this new 
lockdep splat:

[    0.000000] Initializing CPU#0
[    0.000000] found new irq_desc for irq 0
[    0.000000] INFO: trying to register non-static key.
[    0.000000] the code is fine but needs lockdep annotation.
[    0.000000] turning off the locking correctness validator.
[    0.000000] Pid: 0, comm: swapper Not tainted 2.6.27-rc3-tip-00191-g98ccb89-dirty #1
[    0.000000]  [<c0153c22>] register_lock_class+0x3d2/0x400
[    0.000000]  [<c0104d87>] ? mcount_call+0x5/0xa
[    0.000000]  [<c0154f3a>] __lock_acquire+0x22a/0x5d0
[    0.000000]  [<c0104d87>] ? mcount_call+0x5/0xa
[    0.000000]  [<c0155351>] lock_acquire+0x71/0xa0
[    0.000000]  [<c016d61f>] ? set_irq_chip+0x3f/0x90
[    0.000000]  [<c070f148>] _spin_lock_irqsave+0x58/0x90
[    0.000000]  [<c016d61f>] ? set_irq_chip+0x3f/0x90
[    0.000000]  [<c016d61f>] set_irq_chip+0x3f/0x90
[    0.000000]  [<c016d7e0>] ? handle_level_irq+0x0/0xe0
[    0.000000]  [<c016da1a>] set_irq_chip_and_handler_name+0x1a/0x40
[    0.000000]  [<c0a396c1>] init_ISA_irqs+0x51/0xa0
[    0.000000]  [<c0a4a365>] pre_intr_init_hook+0x25/0x30
[    0.000000]  [<c0a39723>] native_init_IRQ+0x13/0x370
[    0.000000]  [<c015569c>] ? lock_release+0xcc/0x1d0
[    0.000000]  [<c0104d87>] ? mcount_call+0x5/0xa
[    0.000000]  [<c070dc22>] ? __mutex_unlock_slowpath+0x92/0x110
[    0.000000]  [<c070dcad>] ? mutex_unlock+0xd/0x10
[    0.000000]  [<c0135f62>] ? cpu_maps_update_done+0x12/0x20
[    0.000000]  [<c06c6743>] ? register_cpu_notifier+0x23/0x30
[    0.000000]  [<c011e8ae>] init_IRQ+0xe/0x10
[    0.000000]  [<c0a357a5>] start_kernel+0x1c5/0x340
[    0.000000]  [<c0a35280>] ? unknown_bootoption+0x0/0x210
[    0.000000]  [<c0a3506b>] i386_start_kernel+0x6b/0x80
[    0.000000]  =======================
[    0.000000] found new irq_desc for irq 1
[    0.000000] found new irq_desc for irq 2
[    0.000000] found new irq_desc for irq 3

this:

 static void init_one_irq_desc(struct irq_desc *desc)
 {
         memcpy(desc, &irq_desc_init, sizeof(struct irq_desc));
 #ifdef CONFIG_TRACE_IRQFLAGS
         lockdep_set_class(&desc->lock, &irq_desc_lock_class);
 #endif
 }

looks wrong - it should be unconditional (i've done that via the patch 
below). I dont remember why we made early_init_irq_lock_class() 
dependent on CONFIG_TRACE_IRQFLAGS before. Peter, do you have any 
memories?

	Ingo

--------------------------->
>From bba3b43f2383c2fd06c4befee45ef30d5cf59d32 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 15 Aug 2008 11:34:35 +0200
Subject: [PATCH] sparseirq: fix lockdep

-tip testing found this lockdep splat:

[    0.000000] Initializing CPU#0
[    0.000000] found new irq_desc for irq 0
[    0.000000] INFO: trying to register non-static key.
[    0.000000] the code is fine but needs lockdep annotation.
[    0.000000] turning off the locking correctness validator.
[    0.000000] Pid: 0, comm: swapper Not tainted 2.6.27-rc3-tip-00191-g98ccb89-dirty #1
[    0.000000]  [<c0153c22>] register_lock_class+0x3d2/0x400
[    0.000000]  [<c0104d87>] ? mcount_call+0x5/0xa
[    0.000000]  [<c0154f3a>] __lock_acquire+0x22a/0x5d0
[    0.000000]  [<c0104d87>] ? mcount_call+0x5/0xa
[    0.000000]  [<c0155351>] lock_acquire+0x71/0xa0
[    0.000000]  [<c016d61f>] ? set_irq_chip+0x3f/0x90
[    0.000000]  [<c070f148>] _spin_lock_irqsave+0x58/0x90
[    0.000000]  [<c016d61f>] ? set_irq_chip+0x3f/0x90
[    0.000000]  [<c016d61f>] set_irq_chip+0x3f/0x90
[    0.000000]  [<c016d7e0>] ? handle_level_irq+0x0/0xe0
[    0.000000]  [<c016da1a>] set_irq_chip_and_handler_name+0x1a/0x40
[    0.000000]  [<c0a396c1>] init_ISA_irqs+0x51/0xa0
[    0.000000]  [<c0a4a365>] pre_intr_init_hook+0x25/0x30
[    0.000000]  [<c0a39723>] native_init_IRQ+0x13/0x370
[    0.000000]  [<c015569c>] ? lock_release+0xcc/0x1d0
[    0.000000]  [<c0104d87>] ? mcount_call+0x5/0xa
[    0.000000]  [<c070dc22>] ? __mutex_unlock_slowpath+0x92/0x110
[    0.000000]  [<c070dcad>] ? mutex_unlock+0xd/0x10
[    0.000000]  [<c0135f62>] ? cpu_maps_update_done+0x12/0x20
[    0.000000]  [<c06c6743>] ? register_cpu_notifier+0x23/0x30
[    0.000000]  [<c011e8ae>] init_IRQ+0xe/0x10
[    0.000000]  [<c0a357a5>] start_kernel+0x1c5/0x340
[    0.000000]  [<c0a35280>] ? unknown_bootoption+0x0/0x210
[    0.000000]  [<c0a3506b>] i386_start_kernel+0x6b/0x80
[    0.000000]  =======================
[    0.000000] found new irq_desc for irq 1
[    0.000000] found new irq_desc for irq 2
[    0.000000] found new irq_desc for irq 3

this:

 static void init_one_irq_desc(struct irq_desc *desc)
 {
         memcpy(desc, &irq_desc_init, sizeof(struct irq_desc));
 #ifdef CONFIG_TRACE_IRQFLAGS
         lockdep_set_class(&desc->lock, &irq_desc_lock_class);
 #endif
 }

should be unconditional.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 kernel/irq/handle.c |    2 --
 1 files changed, 0 insertions(+), 2 deletions(-)

diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c
index ce8774b..8cd0c68 100644
--- a/kernel/irq/handle.c
+++ b/kernel/irq/handle.c
@@ -75,9 +75,7 @@ static struct irq_desc irq_desc_init = {
 static void init_one_irq_desc(struct irq_desc *desc)
 {
 	memcpy(desc, &irq_desc_init, sizeof(struct irq_desc));
-#ifdef CONFIG_TRACE_IRQFLAGS
 	lockdep_set_class(&desc->lock, &irq_desc_lock_class);
-#endif
 }
 
 extern int after_bootmem;

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Peter Zijlstra]

On Fri, 2008-08-15 at 11:35 +0200, Ingo Molnar wrote:

> looks wrong - it should be unconditional (i've done that via the patch 
> below). I dont remember why we made early_init_irq_lock_class() 
> dependent on CONFIG_TRACE_IRQFLAGS before. Peter, do you have any 
> memories?

Nope, git suggests I poke you about it.. :-)

commit 243c7621aac4ed1aa79524c9a1cecf7c05a28124
Author: Ingo Molnar <mingo@elte.hu>
Date:   Mon Jul 3 00:25:06 2006 -0700

    [PATCH] lockdep: annotate genirq
    
    Teach special (recursive) locking code to the lock validator.  Has no effect
    on non-lockdep kernels.
    
    Signed-off-by: Ingo Molnar <mingo@elte.hu>
    Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
    Signed-off-by: Andrew Morton <akpm@osdl.org>
    Signed-off-by: Linus Torvalds <torvalds@osdl.org>

diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h
index 80ec7a4..316e0fb 100644
--- a/include/linux/lockdep.h
+++ b/include/linux/lockdep.h
@@ -270,6 +270,12 @@ static inline int lockdep_internal(void)
 struct lock_class_key { };
 #endif /* !LOCKDEP */
 
+#if defined(CONFIG_TRACE_IRQFLAGS) && defined(CONFIG_GENERIC_HARDIRQS)
+extern void early_init_irq_lock_class(void);
+#else
+# define early_init_irq_lock_class()		do { } while (0)
+#endif
+
 #ifdef CONFIG_TRACE_IRQFLAGS
 extern void early_boot_irqs_off(void);
 extern void early_boot_irqs_on(void);
diff --git a/init/main.c b/init/main.c
index fc473d4..628b8e9 100644
--- a/init/main.c
+++ b/init/main.c
@@ -466,6 +466,7 @@ asmlinkage void __init start_kernel(void)
 
 	local_irq_disable();
 	early_boot_irqs_off();
+	early_init_irq_lock_class();
 
 /*
  * Interrupts are still disabled. Do necessary setups, then
diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c
index aeb6e39..a7b497e 100644
--- a/kernel/irq/handle.c
+++ b/kernel/irq/handle.c
@@ -249,3 +249,19 @@ out:
 	return 1;
 }
 
+#ifdef CONFIG_TRACE_IRQFLAGS
+
+/*
+ * lockdep: we want to handle all irq_desc locks as a single lock-class:
+ */
+static struct lock_class_key irq_desc_lock_class;
+
+void early_init_irq_lock_class(void)
+{
+	int i;
+
+	for (i = 0; i < NR_IRQS; i++)
+		lockdep_set_class(&irq_desc[i].lock, &irq_desc_lock_class);
+}
+
+#endif

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Ingo Molnar]

* Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:

> On Fri, 2008-08-15 at 11:35 +0200, Ingo Molnar wrote:
> 
> > looks wrong - it should be unconditional (i've done that via the patch 
> > below). I dont remember why we made early_init_irq_lock_class() 
> > dependent on CONFIG_TRACE_IRQFLAGS before. Peter, do you have any 
> > memories?
> 
> Nope, git suggests I poke you about it.. :-)

muhaha - i dont even know that guy who did this sloppy commit 2 years 
ago - "Ingo Molnar", or however he is called.

I believe my (mistaken) thinking would have been that desc->irq lockdep 
class does not matter as long as we dont do irq-tracing. Which might 
have been borderline correct then but is wrong now, for things like 
lockstat?

	Ingo

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Peter Zijlstra]

On Fri, 2008-08-15 at 12:19 +0200, Ingo Molnar wrote:
> * Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:
> 
> > On Fri, 2008-08-15 at 11:35 +0200, Ingo Molnar wrote:
> > 
> > > looks wrong - it should be unconditional (i've done that via the patch 
> > > below). I dont remember why we made early_init_irq_lock_class() 
> > > dependent on CONFIG_TRACE_IRQFLAGS before. Peter, do you have any 
> > > memories?
> > 
> > Nope, git suggests I poke you about it.. :-)
> 
> muhaha - i dont even know that guy who did this sloppy commit 2 years 
> ago - "Ingo Molnar", or however he is called.
> 
> I believe my (mistaken) thinking would have been that desc->irq lockdep 
> class does not matter as long as we dont do irq-tracing. Which might 
> have been borderline correct then but is wrong now, for things like 
> lockstat?

/me pokes a bit at kconfig dependancies, and yes, it seems you're right
- lockstat doesn't require irq tracing..

Re: [PATCH 00/07] dyn_array/nr_irqs/sparse_irq support v10 - fix

[Yinghai Lu]

On Fri, Aug 15, 2008 at 3:28 AM, Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:
> On Fri, 2008-08-15 at 12:19 +0200, Ingo Molnar wrote:
>> * Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:
>>
>> > On Fri, 2008-08-15 at 11:35 +0200, Ingo Molnar wrote:
>> >
>> > > looks wrong - it should be unconditional (i've done that via the patch
>> > > below). I dont remember why we made early_init_irq_lock_class()
>> > > dependent on CONFIG_TRACE_IRQFLAGS before. Peter, do you have any
>> > > memories?
>> >
>> > Nope, git suggests I poke you about it.. :-)
>>
>> muhaha - i dont even know that guy who did this sloppy commit 2 years
>> ago - "Ingo Molnar", or however he is called.
>>
>> I believe my (mistaken) thinking would have been that desc->irq lockdep
>> class does not matter as long as we dont do irq-tracing. Which might
>> have been borderline correct then but is wrong now, for things like
>> lockstat?
>
> /me pokes a bit at kconfig dependancies, and yes, it seems you're right
> - lockstat doesn't require irq tracing..

can you check if

#ifdef CONFIG_TRACE_IRQFLAGS
void early_init_irq_lock_class(void)
{
#ifndef CONFIG_HAVE_DYN_ARRAY
        int i;

        for (i = 0; i < nr_irqs; i++)
                lockdep_set_class(&irq_descX[i].lock, &irq_desc_lock_class);
#endif
}
#endif

in kernel/irq/handle.c is right too?

YH

[PATCH 0/7] merge io_apic_xx.c

[Yinghai Lu]

    merge io_apic_xx.c

Thanks

Yinghai Lu

[PATCH 1/7] x86: ordering functions in io_apic_32.c - fix

[Yinghai Lu]

add calling setup_msi_irq back to arch_setup_msi_irq

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

Index: linux-2.6/arch/x86/kernel/io_apic_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_32.c
+++ linux-2.6/arch/x86/kernel/io_apic_32.c
@@ -2686,7 +2695,6 @@ static unsigned int build_irq_for_pci_de
 
 int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
 {
-	struct msi_msg msg;
 	int irq, ret;
         unsigned int irq_want;
 
@@ -2697,6 +2705,7 @@ int arch_setup_msi_irq(struct pci_dev *d
 	if (irq == 0)
 		return -1;
 
+	ret = setup_msi_irq(dev, desc, irq);
 	if (ret < 0) {
 		destroy_irq(irq);
 		return ret;

[PATCH 2/7] x86: make headers files the smae in io_apic_xx.c

[Yinghai Lu]

also make no_timer_check to be global on 64 bit, because vmi_32 is using that.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/io_apic_32.c |   15 ++++++++++++---
 arch/x86/kernel/io_apic_64.c |   12 +++++++++---
 2 files changed, 21 insertions(+), 6 deletions(-)

Index: linux-2.6/arch/x86/kernel/io_apic_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_32.c
+++ linux-2.6/arch/x86/kernel/io_apic_32.c
@@ -25,28 +25,37 @@
 #include <linux/init.h>
 #include <linux/delay.h>
 #include <linux/sched.h>
-#include <linux/bootmem.h>
+#include <linux/pci.h>
 #include <linux/mc146818rtc.h>
 #include <linux/compiler.h>
 #include <linux/acpi.h>
 #include <linux/module.h>
 #include <linux/sysdev.h>
-#include <linux/pci.h>
 #include <linux/msi.h>
 #include <linux/htirq.h>
 #include <linux/freezer.h>
 #include <linux/kthread.h>
-#include <linux/jiffies.h>	/* time_after() */
+#include <linux/jiffies.h>      /* time_after() */
+#ifdef CONFIG_ACPI
+#include <acpi/acpi_bus.h>
+#endif
+#include <linux/bootmem.h>
+#include <linux/dmar.h>
 
+#include <asm/idle.h>
 #include <asm/io.h>
 #include <asm/smp.h>
 #include <asm/desc.h>
+#include <asm/proto.h>
+#include <asm/acpi.h>
+#include <asm/dma.h>
 #include <asm/timer.h>
 #include <asm/i8259.h>
 #include <asm/nmi.h>
 #include <asm/msidef.h>
 #include <asm/hypertransport.h>
 #include <asm/setup.h>
+#include <asm/irq_remapping.h>
 
 #include <mach_ipi.h>
 #include <mach_apic.h>
Index: linux-2.6/arch/x86/kernel/io_apic_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_64.c
+++ linux-2.6/arch/x86/kernel/io_apic_64.c
@@ -27,12 +27,15 @@
 #include <linux/sched.h>
 #include <linux/pci.h>
 #include <linux/mc146818rtc.h>
+#include <linux/compiler.h>
 #include <linux/acpi.h>
+#include <linux/module.h>
 #include <linux/sysdev.h>
 #include <linux/msi.h>
 #include <linux/htirq.h>
-#include <linux/dmar.h>
-#include <linux/jiffies.h>
+#include <linux/freezer.h>
+#include <linux/kthread.h>
+#include <linux/jiffies.h>	/* time_after() */
 #ifdef CONFIG_ACPI
 #include <acpi/acpi_bus.h>
 #endif
@@ -46,14 +49,17 @@
 #include <asm/proto.h>
 #include <asm/acpi.h>
 #include <asm/dma.h>
+#include <asm/timer.h>
 #include <asm/i8259.h>
 #include <asm/nmi.h>
 #include <asm/msidef.h>
 #include <asm/hypertransport.h>
+#include <asm/setup.h>
 #include <asm/irq_remapping.h>
 
 #include <mach_ipi.h>
 #include <mach_apic.h>
+#include <mach_apicdef.h>
 
 #define __apicdebuginit(type) static type __init
 
@@ -1683,7 +1689,7 @@ void disable_IO_APIC(void)
 	disconnect_bsp_APIC(ioapic_i8259.pin != -1);
 }
 
-static int no_timer_check;
+int no_timer_check __initdata;
 
 static int __init notimercheck(char *s)
 {

[PATCH 3/7] x86: make 64 handle sis_apic_bug like the 32 bit

[Yinghai Lu]

do we have 64bit system with sis chipset?

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/io_apic_64.c |   28 ++++++++++++++++++++++++----
 1 file changed, 24 insertions(+), 4 deletions(-)

Index: linux-2.6/arch/x86/kernel/io_apic_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_64.c
+++ linux-2.6/arch/x86/kernel/io_apic_64.c
@@ -65,7 +65,11 @@
 
 int ioapic_force;
 
-int sis_apic_bug; /* not actually supported, dummy for compile */
+/*
+ *      Is the SiS APIC rmw bug present ?
+ *      -1 = don't know, 0 = no, 1 = yes
+ */
+int sis_apic_bug = -1;
 
 static DEFINE_SPINLOCK(ioapic_lock);
 static DEFINE_SPINLOCK(vector_lock);
@@ -373,9 +377,11 @@ static inline void io_apic_write(unsigne
  * 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)
+static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
 {
 	struct io_apic __iomem *io_apic = io_apic_base(apic);
+        if (sis_apic_bug)
+                writel(reg, &io_apic->index);
 	writel(value, &io_apic->data);
 }
 
@@ -494,7 +500,7 @@ static void __target_IO_APIC_irq(unsigne
 		reg = io_apic_read(apic, 0x10 + pin*2);
 		reg &= ~IO_APIC_REDIR_VECTOR_MASK;
 		reg |= vector;
-		io_apic_modify(apic, reg);
+		io_apic_modify(apic, 0x10 + pin*2, reg);
 		if (!entry->next)
 			break;
 		entry = entry->next;
@@ -622,7 +628,7 @@ static inline void io_apic_sync(unsigned
 		pin = entry->pin;					\
 		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
 		reg ACTION;						\
-		io_apic_modify(entry->apic, reg);			\
+		io_apic_modify(entry->apic, 0x10 + R + pin*2, reg);	\
 		FINAL;							\
 		if (!entry->next)					\
 			break;						\
@@ -2446,6 +2452,20 @@ void __init setup_IO_APIC(void)
 	check_timer();
 }
 
+/*
+ *      Called after all the initialization is done. If we didnt find any
+ *      APIC bugs then we can allow the modify fast path
+ */
+
+static int __init io_apic_bug_finalize(void)
+{
+        if (sis_apic_bug == -1)
+                sis_apic_bug = 0;
+        return 0;
+}
+
+late_initcall(io_apic_bug_finalize);
+
 struct sysfs_ioapic_data {
 	struct sys_device dev;
 	struct IO_APIC_route_entry entry[0];

[PATCH 4/7] x86: remve ioapic_force

[Yinghai Lu]

no user

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c
index 6709393..cf89524 100644
--- a/arch/x86/kernel/apic_64.c
+++ b/arch/x86/kernel/apic_64.c
@@ -1555,7 +1555,6 @@ static int __init apic_set_verbosity(char *str)
 {
 	if (str == NULL)  {
 		skip_ioapic_setup = 0;
-		ioapic_force = 1;
 		return 0;
 	}
 	if (strcmp("debug", str) == 0)
diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c
index bb024d1..ca87077 100644
--- a/arch/x86/kernel/io_apic_64.c
+++ b/arch/x86/kernel/io_apic_64.c
@@ -63,8 +63,6 @@
 
 #define __apicdebuginit(type) static type __init
 
-int ioapic_force;
-
 /*
  *      Is the SiS APIC rmw bug present ?
  *      -1 = don't know, 0 = no, 1 = yes
diff --git a/include/asm-x86/apic.h b/include/asm-x86/apic.h
index d76a083..2d970f6 100644
--- a/include/asm-x86/apic.h
+++ b/include/asm-x86/apic.h
@@ -40,8 +40,6 @@ extern void generic_apic_probe(void);
 extern unsigned int apic_verbosity;
 extern int local_apic_timer_c2_ok;
 
-extern int ioapic_force;
-
 extern int disable_apic;
 /*
  * Basic functions accessing APICs.

[PATCH 5/7] x86: make io_apic_64.c and io_apic_32.c the same

[Yinghai Lu]

all the same except INTR_REMAPPING related and ioapic io resource

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/io_apic_32.c |  213 ++++++++++++-
 arch/x86/kernel/io_apic_64.c |  666 ++++++++++++++++++++++++++++++++++++++++---
 2 files changed, 832 insertions(+), 47 deletions(-)

Index: linux-2.6/arch/x86/kernel/io_apic_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_32.c
+++ linux-2.6/arch/x86/kernel/io_apic_32.c
@@ -123,7 +123,6 @@ struct irq_cfg {
 	u8 move_in_progress : 1;
 };
 
-
 /* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
 static struct irq_cfg irq_cfg_legacy[] __initdata = {
 	[0]  = { .irq =  0, .domain = CPU_MASK_ALL, .vector = IRQ0_VECTOR,  },
@@ -391,6 +390,38 @@ static inline void io_apic_modify(unsign
 	writel(value, &io_apic->data);
 }
 
+#ifdef CONFIG_X86_64
+static bool io_apic_level_ack_pending(unsigned int irq)
+{
+	struct irq_pin_list *entry;
+	unsigned long flags;
+	struct irq_cfg *cfg = irq_cfg(irq);
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	entry = cfg->irq_2_pin;
+	for (;;) {
+		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? */
+		if (reg & IO_APIC_REDIR_REMOTE_IRR) {
+			spin_unlock_irqrestore(&ioapic_lock, flags);
+			return true;
+		}
+		if (!entry->next)
+			break;
+		entry = entry->next;
+	}
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	return false;
+}
+#endif
+
 union entry_union {
 	struct { u32 w1, w2; };
 	struct IO_APIC_route_entry entry;
@@ -483,17 +514,15 @@ static void set_ioapic_affinity_irq(unsi
 	unsigned int dest;
 	cpumask_t tmp;
 
-	cfg = irq_cfg(irq);
-
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
 		return;
 
+	cfg = irq_cfg(irq);
 	if (assign_irq_vector(irq, mask))
 		return;
 
 	cpus_and(tmp, cfg->domain, mask);
-
 	dest = cpu_mask_to_apicid(tmp);
 	/*
 	 * Only the high 8 bits are valid.
@@ -572,6 +601,54 @@ static void __init replace_pin_at_irq(un
 		add_pin_to_irq(irq, newapic, newpin);
 }
 
+#ifdef CONFIG_X86_64
+/*
+ * Synchronize the IO-APIC and the CPU by doing
+ * a dummy read from the IO-APIC
+ */
+static inline void io_apic_sync(unsigned int apic)
+{
+	struct io_apic __iomem *io_apic = io_apic_base(apic);
+	readl(&io_apic->data);
+}
+
+#define __DO_ACTION(R, ACTION, FINAL)					\
+									\
+{									\
+	int pin;							\
+	struct irq_cfg *cfg;						\
+	struct irq_pin_list *entry;					\
+									\
+	cfg = irq_cfg(irq);						\
+	entry = cfg->irq_2_pin;						\
+	for (;;) {							\
+		unsigned int reg;					\
+		if (!entry)						\
+			break;						\
+		pin = entry->pin;					\
+		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
+		reg ACTION;						\
+		io_apic_modify(entry->apic, 0x10 + R + pin*2, reg);	\
+		FINAL;							\
+		if (!entry->next)					\
+			break;						\
+		entry = entry->next;					\
+	}								\
+}
+
+#define DO_ACTION(name,R,ACTION, FINAL)					\
+									\
+	static void name##_IO_APIC_irq (unsigned int irq)		\
+	__DO_ACTION(R, ACTION, FINAL)
+
+/* mask = 1 */
+DO_ACTION(__mask,	0, |= IO_APIC_REDIR_MASKED, io_apic_sync(entry->apic))
+
+/* mask = 0 */
+DO_ACTION(__unmask,	0, &= ~IO_APIC_REDIR_MASKED, )
+
+#else
+
 static void __modify_IO_APIC_irq(unsigned int irq, unsigned long enable, unsigned long disable)
 {
 	struct irq_cfg *cfg;
@@ -620,6 +697,8 @@ static void __unmask_and_level_IO_APIC_i
 				IO_APIC_REDIR_MASKED);
 }
 
+#endif
+
 static void mask_IO_APIC_irq(unsigned int irq)
 {
 	unsigned long flags;
@@ -1055,6 +1134,17 @@ void unlock_vector_lock(void)
 
 static int __assign_irq_vector(int irq, cpumask_t mask)
 {
+	/*
+	 * NOTE! The local APIC isn't very good at handling
+	 * multiple interrupts at the same interrupt level.
+	 * As the interrupt level is determined by taking the
+	 * vector number and shifting that right by 4, we
+	 * want to spread these out a bit so that they don't
+	 * all fall in the same interrupt level.
+	 *
+	 * Also, we've got to be careful not to trash gate
+	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
+	 */
         static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0;
         unsigned int old_vector;
         int cpu;
@@ -1095,9 +1185,13 @@ next:
                 }
                 if (unlikely(current_vector == vector))
                         continue;
-		if (vector == SYSCALL_VECTOR)
+#ifdef CONFIG_X86_64
+                if (vector == IA32_SYSCALL_VECTOR)
                         goto next;
-
+#else
+                if (vector == SYSCALL_VECTOR)
+                        goto next;
+#endif
                 for_each_cpu_mask_nr(new_cpu, new_mask)
                         if (per_cpu(vector_irq, new_cpu)[vector] != -1)
                                 goto next;
@@ -1184,6 +1278,7 @@ static struct irq_chip ioapic_chip;
 #define IOAPIC_EDGE	0
 #define IOAPIC_LEVEL	1
 
+#ifdef CONFIG_X86_32
 static inline int IO_APIC_irq_trigger(int irq)
 {
 	int apic, idx, pin;
@@ -1200,6 +1295,12 @@ static inline int IO_APIC_irq_trigger(in
 	 */
 	return 0;
 }
+#else
+static inline int IO_APIC_irq_trigger(int irq)
+{
+        return 1;
+}
+#endif
 
 static void ioapic_register_intr(int irq, unsigned long trigger)
 {
@@ -1212,15 +1313,18 @@ static void ioapic_register_intr(int irq
 		desc = irq_desc_with_new(irq);
 
 	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-	    trigger == IOAPIC_LEVEL) {
+	    trigger == IOAPIC_LEVEL)
 		desc->status |= IRQ_LEVEL;
+	else
+		desc->status &= ~IRQ_LEVEL;
+
+	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
+	    trigger == IOAPIC_LEVEL)
 		set_irq_chip_and_handler_name(irq, &ioapic_chip,
 					 handle_fasteoi_irq, "fasteoi");
-	} else {
-		desc->status &= ~IRQ_LEVEL;
+	else
 		set_irq_chip_and_handler_name(irq, &ioapic_chip,
 					 handle_edge_irq, "edge");
-	}
 }
 
 static int setup_ioapic_entry(int apic, int irq,
@@ -1662,7 +1766,6 @@ static void __init enable_IO_APIC(void)
 			struct IO_APIC_route_entry entry;
 			entry = ioapic_read_entry(apic, pin);
 
-
 			/* If the interrupt line is enabled and in ExtInt mode
 			 * I have found the pin where the i8259 is connected.
 			 */
@@ -2012,6 +2115,60 @@ static void ack_apic_edge(unsigned int i
 	ack_APIC_irq();
 }
 
+#ifdef CONFIG_X86_64
+static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
+{
+        int do_unmask_irq = 0;
+
+        irq_complete_move(irq);
+#ifdef CONFIG_GENERIC_PENDING_IRQ
+        /* If we are moving the irq we need to mask it */
+        if (unlikely(desc->status & IRQ_MOVE_PENDING)) {
+                do_unmask_irq = 1;
+                mask_IO_APIC_irq(irq);
+        }
+#endif
+
+        /*
+         * We must acknowledge the irq before we move it or the acknowledge will
+         * not propagate properly.
+         */
+        ack_APIC_irq();
+
+        /* Now we can move and renable the irq */
+        if (unlikely(do_unmask_irq)) {
+                /* Only migrate the irq if the ack has been received.
+                 *
+                 * On rare occasions the broadcast level triggered ack gets
+                 * delayed going to ioapics, and if we reprogram the
+                 * vector while Remote IRR is still set the irq will never
+                 * fire again.
+                 *
+                 * To prevent this scenario we read the Remote IRR bit
+                 * of the ioapic.  This has two effects.
+                 * - On any sane system the read of the ioapic will
+                 *   flush writes (and acks) going to the ioapic from
+                 *   this cpu.
+                 * - We get to see if the ACK has actually been delivered.
+                 *
+                 * Based on failed experiments of reprogramming the
+                 * ioapic entry from outside of irq context starting
+                 * with masking the ioapic entry and then polling until
+                 * Remote IRR was clear before reprogramming the
+                 * ioapic I don't trust the Remote IRR bit to be
+                 * completey accurate.
+                 *
+                 * However there appears to be no other way to plug
+                 * this race, so if the Remote IRR bit is not
+                 * accurate and is causing problems then it is a hardware bug
+                 * and you can go talk to the chipset vendor about it.
+                 */
+                if (!io_apic_level_ack_pending(irq))
+                        move_masked_irq(irq, desc);
+                unmask_IO_APIC_irq(irq);
+        }
+}
+#else
 atomic_t irq_mis_count;
 static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
 {
@@ -2053,6 +2210,7 @@ static void ack_apic_level(unsigned int
 		spin_unlock(&ioapic_lock);
 	}
 }
+#endif
 
 static struct irq_chip ioapic_chip __read_mostly = {
 	.name 		= "IO-APIC",
@@ -2224,7 +2382,7 @@ static inline void __init unlock_ExtINT_
 }
 
 static int disable_timer_pin_1 __initdata;
-
+/* Actually the next is obsolete, but keep it for paranoid reasons -AK */
 static int __init parse_disable_timer_pin_1(char *arg)
 {
 	disable_timer_pin_1 = 1;
@@ -2244,9 +2402,9 @@ static inline void __init check_timer(vo
 {
 	struct irq_cfg *cfg = irq_cfg(0);
 	int apic1, pin1, apic2, pin2;
-	int no_pin1 = 0;
-	unsigned int ver;
 	unsigned long flags;
+	unsigned int ver;
+	int no_pin1 = 0;
 
 	local_irq_save(flags);
 
@@ -2550,6 +2708,7 @@ unsigned int create_irq(unsigned int irq
 		cfg_new = irq_cfg(new);
 		if (cfg_new && cfg_new->vector != 0)
 			continue;
+		/* check if need to create one */
 		if (!cfg_new)
 			cfg_new = irq_cfg_with_new(new);
 		if (__assign_irq_vector(new, TARGET_CPUS) == 0)
@@ -2714,6 +2873,32 @@ int arch_setup_msi_irq(struct pci_dev *d
 	return 0;
 }
 
+int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
+{
+        unsigned int irq;
+        int ret, sub_handle;
+        struct msi_desc *desc;
+        unsigned int irq_want;
+
+        irq_want = build_irq_for_pci_dev(dev) + 0x100;
+        sub_handle = 0;
+        list_for_each_entry(desc, &dev->msi_list, list) {
+                irq = create_irq(irq_want--);
+                if (irq == 0)
+                        return -1;
+                ret = setup_msi_irq(dev, desc, irq);
+                if (ret < 0)
+                        goto error;
+                sub_handle++;
+        }
+        return 0;
+
+error:
+        destroy_irq(irq);
+        return ret;
+}
+
+
 void arch_teardown_msi_irq(unsigned int irq)
 {
 	destroy_irq(irq);
Index: linux-2.6/arch/x86/kernel/io_apic_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_64.c
+++ linux-2.6/arch/x86/kernel/io_apic_64.c
@@ -94,18 +94,22 @@ struct mp_config_intsrc mp_irqs[MAX_IRQ_
 /* # of MP IRQ source entries */
 int mp_irq_entries;
 
+#if defined (CONFIG_MCA) || defined (CONFIG_EISA)
+int mp_bus_id_to_type[MAX_MP_BUSSES];
+#endif
+
 DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES);
 
 int skip_ioapic_setup;
 
 static int __init parse_noapic(char *str)
 {
+	/* disable IO-APIC */
 	disable_ioapic_setup();
 	return 0;
 }
 early_param("noapic", parse_noapic);
 
-
 struct irq_cfg;
 struct irq_pin_list;
 struct irq_cfg {
@@ -374,6 +378,8 @@ static inline void io_apic_write(unsigne
 /*
  * Re-write a value: to be used for read-modify-write
  * cycles where the read already set up the index register.
+ *
+ * Older SiS APIC requires we rewrite the index register
  */
 static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
 {
@@ -383,6 +389,7 @@ static inline void io_apic_modify(unsign
 	writel(value, &io_apic->data);
 }
 
+#ifdef CONFIG_X86_64
 static bool io_apic_level_ack_pending(unsigned int irq)
 {
 	struct irq_pin_list *entry;
@@ -412,6 +419,7 @@ static bool io_apic_level_ack_pending(un
 
 	return false;
 }
+#endif
 
 union entry_union {
 	struct { u32 w1, w2; };
@@ -509,7 +517,7 @@ static int assign_irq_vector(int irq, cp
 
 static void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
 {
-	struct irq_cfg *cfg = irq_cfg(irq);
+	struct irq_cfg *cfg;
 	unsigned long flags;
 	unsigned int dest;
 	cpumask_t tmp;
@@ -518,12 +526,12 @@ static void set_ioapic_affinity_irq(unsi
 	if (cpus_empty(tmp))
 		return;
 
+	cfg = irq_cfg(irq);
 	if (assign_irq_vector(irq, mask))
 		return;
 
 	cpus_and(tmp, cfg->domain, mask);
 	dest = cpu_mask_to_apicid(tmp);
-
 	/*
 	 * Only the high 8 bits are valid.
 	 */
@@ -534,7 +542,7 @@ static void set_ioapic_affinity_irq(unsi
 	desc->affinity = mask;
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 }
-#endif
+#endif /* CONFIG_SMP */
 
 /*
  * The common case is 1:1 IRQ<->pin mappings. Sometimes there are
@@ -600,6 +608,7 @@ static void __init replace_pin_at_irq(un
 		add_pin_to_irq(irq, newapic, newpin);
 }
 
+#ifdef CONFIG_X86_64
 /*
  * Synchronize the IO-APIC and the CPU by doing
  * a dummy read from the IO-APIC
@@ -645,6 +654,58 @@ DO_ACTION(__mask,	0, |= IO_APIC_REDIR_MA
 /* mask = 0 */
 DO_ACTION(__unmask,	0, &= ~IO_APIC_REDIR_MASKED, )
 
+#else
+
+static void __modify_IO_APIC_irq(unsigned int irq, unsigned long enable, unsigned long disable)
+{
+	struct irq_cfg *cfg;
+	struct irq_pin_list *entry;
+	unsigned int pin, reg;
+
+	cfg = irq_cfg(irq);
+	entry = cfg->irq_2_pin;
+	for (;;) {
+		if (!entry)
+			break;
+		pin = entry->pin;
+		reg = io_apic_read(entry->apic, 0x10 + pin*2);
+		reg &= ~disable;
+		reg |= enable;
+		io_apic_modify(entry->apic, 0x10 + pin*2, reg);
+		if (!entry->next)
+			break;
+		entry = entry->next;
+	}
+}
+
+/* mask = 1 */
+static void __mask_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED, 0);
+}
+
+/* mask = 0 */
+static void __unmask_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, 0, IO_APIC_REDIR_MASKED);
+}
+
+/* mask = 1, trigger = 0 */
+static void __mask_and_edge_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED,
+				IO_APIC_REDIR_LEVEL_TRIGGER);
+}
+
+/* mask = 0, trigger = 1 */
+static void __unmask_and_level_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_LEVEL_TRIGGER,
+				IO_APIC_REDIR_MASKED);
+}
+
+#endif
+
 static void mask_IO_APIC_irq (unsigned int irq)
 {
 	unsigned long flags;
@@ -686,6 +747,64 @@ static void clear_IO_APIC (void)
 			clear_IO_APIC_pin(apic, pin);
 }
 
+#if !defined(CONFIG_SMP) && defined(CONFIG_X86_32)
+void send_IPI_self(int vector)
+{
+	unsigned int cfg;
+
+	/*
+	 * Wait for idle.
+	 */
+	apic_wait_icr_idle();
+	cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
+	/*
+	 * Send the IPI. The write to APIC_ICR fires this off.
+	 */
+	apic_write(APIC_ICR, cfg);
+}
+#endif /* !CONFIG_SMP && CONFIG_X86_32*/
+
+#ifdef CONFIG_X86_32
+/*
+ * 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;
+
+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, KERN_INFO
+			"PIRQ redirection, working around broken MP-BIOS.\n");
+	max = MAX_PIRQS;
+	if (ints[0] < MAX_PIRQS)
+		max = ints[0];
+
+	for (i = 0; i < max; i++) {
+		apic_printk(APIC_VERBOSE, KERN_DEBUG
+				"... PIRQ%d -> IRQ %d\n", i, ints[i+1]);
+		/*
+		 * PIRQs are mapped upside down, usually.
+		 */
+		pirq_entries[MAX_PIRQS-i-1] = ints[i+1];
+	}
+	return 1;
+}
+
+__setup("pirq=", ioapic_pirq_setup);
+#endif /* CONFIG_X86_32 */
+
 #ifdef CONFIG_INTR_REMAP
 /* I/O APIC RTE contents at the OS boot up */
 static struct IO_APIC_route_entry *early_ioapic_entries[MAX_IO_APICS];
@@ -859,18 +978,54 @@ int IO_APIC_get_PCI_irq_vector(int bus,
 	return best_guess;
 }
 
+EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
+
+#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
+/*
+ * 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, KERN_INFO
+			"Broken MPtable reports ISA irq %d\n", irq);
+	return 0;
+}
+
+#endif
+
 /* ISA interrupts are always polarity zero edge triggered,
  * when listed as conforming in the MP table. */
 
 #define default_ISA_trigger(idx)	(0)
 #define default_ISA_polarity(idx)	(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].mp_srcbusirq))
+#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
+
 /* PCI interrupts are always polarity one level triggered,
  * when listed as conforming in the MP table. */
 
 #define default_PCI_trigger(idx)	(1)
 #define default_PCI_polarity(idx)	(1)
 
+/* 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].mp_srcbusirq))
+#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
+
+
 static int MPBIOS_polarity(int idx)
 {
 	int bus = mp_irqs[idx].mp_srcbus;
@@ -928,6 +1083,36 @@ static int MPBIOS_trigger(int idx)
 				trigger = default_ISA_trigger(idx);
 			else
 				trigger = default_PCI_trigger(idx);
+#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
+			switch (mp_bus_id_to_type[bus]) {
+				case MP_BUS_ISA: /* ISA pin */
+				{
+					/* set before the switch */
+					break;
+				}
+				case MP_BUS_EISA: /* EISA pin */
+				{
+					trigger = default_EISA_trigger(idx);
+					break;
+				}
+				case MP_BUS_PCI: /* PCI pin */
+				{
+					/* set before the switch */
+					break;
+				}
+				case MP_BUS_MCA: /* MCA pin */
+				{
+					trigger = default_MCA_trigger(idx);
+					break;
+				}
+				default:
+				{
+					printk(KERN_WARNING "broken BIOS!!\n");
+					trigger = 1;
+					break;
+				}
+			}
+#endif
 			break;
 		case 1: /* edge */
 		{
@@ -965,6 +1150,7 @@ static inline int irq_trigger(int idx)
 	return MPBIOS_trigger(idx);
 }
 
+int (*ioapic_renumber_irq)(int ioapic, int irq);
 static int pin_2_irq(int idx, int apic, int pin)
 {
 	int irq, i;
@@ -986,7 +1172,32 @@ static int pin_2_irq(int idx, int apic,
 		while (i < apic)
 			irq += nr_ioapic_registers[i++];
 		irq += pin;
+                /*
+                 * For MPS mode, so far only needed by ES7000 platform
+                 */
+                if (ioapic_renumber_irq)
+                        irq = ioapic_renumber_irq(apic, irq);
+	}
+
+#ifdef CONFIG_X86_32
+	/*
+	 * PCI IRQ command line redirection. Yes, limits are hardcoded.
+	 */
+	if ((pin >= 16) && (pin <= 23)) {
+		if (pirq_entries[pin-16] != -1) {
+			if (!pirq_entries[pin-16]) {
+				apic_printk(APIC_VERBOSE, KERN_DEBUG
+						"disabling PIRQ%d\n", pin-16);
+			} else {
+				irq = pirq_entries[pin-16];
+				apic_printk(APIC_VERBOSE, KERN_DEBUG
+						"using PIRQ%d -> IRQ %d\n",
+						pin-16, irq);
+			}
+		}
 	}
+#endif
+
 	return irq;
 }
 
@@ -1056,8 +1267,13 @@ next:
 		}
 		if (unlikely(current_vector == vector))
 			continue;
+#ifdef CONFIG_X86_64
 		if (vector == IA32_SYSCALL_VECTOR)
 			goto next;
+#else
+		if (vector == SYSCALL_VECTOR)
+			goto next;
+#endif
 		for_each_cpu_mask_nr(new_cpu, new_mask)
 			if (per_cpu(vector_irq, new_cpu)[vector] != -1)
 				goto next;
@@ -1138,6 +1354,34 @@ static struct irq_chip ioapic_chip;
 static struct irq_chip ir_ioapic_chip;
 #endif
 
+#define IOAPIC_AUTO     -1
+#define IOAPIC_EDGE     0
+#define IOAPIC_LEVEL    1
+
+#ifdef CONFIG_X86_32
+static inline int IO_APIC_irq_trigger(int irq)
+{
+        int apic, idx, pin;
+
+        for (apic = 0; apic < nr_ioapics; apic++) {
+                for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+                        idx = find_irq_entry(apic, pin, mp_INT);
+                        if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
+                                return irq_trigger(idx);
+                }
+        }
+        /*
+         * nonexistent IRQs are edge default
+         */
+        return 0;
+}
+#else
+static inline int IO_APIC_irq_trigger(int irq)
+{
+	return 1;
+}
+#endif
+
 static void ioapic_register_intr(int irq, unsigned long trigger)
 {
 	struct irq_desc *desc;
@@ -1148,7 +1392,8 @@ static void ioapic_register_intr(int irq
 	else
 		desc = irq_desc_with_new(irq);
 
-	if (trigger)
+	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
+	    trigger == IOAPIC_LEVEL)
 		desc->status |= IRQ_LEVEL;
 	else
 		desc->status &= ~IRQ_LEVEL;
@@ -1166,7 +1411,8 @@ static void ioapic_register_intr(int irq
 		return;
 	}
 #endif
-	if (trigger)
+	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
+	    trigger == IOAPIC_LEVEL)
 		set_irq_chip_and_handler_name(irq, &ioapic_chip,
 					      handle_fasteoi_irq,
 					      "fasteoi");
@@ -1301,6 +1547,10 @@ static void __init setup_IO_APIC_irqs(vo
 		}
 
 		irq = pin_2_irq(idx, apic, pin);
+#ifdef CONFIG_X86_32
+                if (multi_timer_check(apic, irq))
+                        continue;
+#endif
 		add_pin_to_irq(irq, apic, pin);
 
 		setup_IO_APIC_irq(apic, pin, irq,
@@ -1358,6 +1608,7 @@ __apicdebuginit(void) print_IO_APIC(void
 	union IO_APIC_reg_00 reg_00;
 	union IO_APIC_reg_01 reg_01;
 	union IO_APIC_reg_02 reg_02;
+	union IO_APIC_reg_03 reg_03;
 	unsigned long flags;
 	struct irq_cfg *cfg;
 
@@ -1382,6 +1633,8 @@ __apicdebuginit(void) print_IO_APIC(void
 	reg_01.raw = io_apic_read(apic, 1);
 	if (reg_01.bits.version >= 0x10)
 		reg_02.raw = io_apic_read(apic, 2);
+        if (reg_01.bits.version >= 0x20)
+                reg_03.raw = io_apic_read(apic, 3);
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 
 	printk("\n");
@@ -1397,11 +1650,27 @@ __apicdebuginit(void) print_IO_APIC(void
 	printk(KERN_DEBUG ".......     : PRQ implemented: %X\n", reg_01.bits.PRQ);
 	printk(KERN_DEBUG ".......     : IO APIC version: %04X\n", reg_01.bits.version);
 
-	if (reg_01.bits.version >= 0x10) {
+	/*
+	 * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02,
+	 * but the value of reg_02 is read as the previous read register
+	 * value, so ignore it if reg_02 == reg_01.
+	 */
+	if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) {
 		printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
 		printk(KERN_DEBUG ".......     : arbitration: %02X\n", reg_02.bits.arbitration);
 	}
 
+	/*
+	 * Some Intel chipsets with IO APIC VERSION of 0x2? don't have reg_02
+	 * or reg_03, but the value of reg_0[23] is read as the previous read
+	 * register value, so ignore it if reg_03 == reg_0[12].
+	 */
+	if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw &&
+	    reg_03.raw != reg_01.raw) {
+		printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw);
+		printk(KERN_DEBUG ".......     : Boot DT    : %X\n", reg_03.bits.boot_DT);
+	}
+
 	printk(KERN_DEBUG ".... IRQ redirection table:\n");
 
 	printk(KERN_DEBUG " NR Dst Mask Trig IRR Pol"
@@ -1473,7 +1742,7 @@ __apicdebuginit(void) print_APIC_bitfiel
 __apicdebuginit(void) print_local_APIC(void *dummy)
 {
 	unsigned int v, ver, maxlvt;
-	unsigned long icr;
+	u64 icr;
 
 	if (apic_verbosity == APIC_QUIET)
 		return;
@@ -1490,11 +1759,13 @@ __apicdebuginit(void) print_local_APIC(v
 	v = apic_read(APIC_TASKPRI);
 	printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
 
-	v = apic_read(APIC_ARBPRI);
-	printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
-		v & APIC_ARBPRI_MASK);
-	v = apic_read(APIC_PROCPRI);
-	printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
+	if (APIC_INTEGRATED(ver)) {                     /* !82489DX */
+		v = apic_read(APIC_ARBPRI);
+		printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
+			v & APIC_ARBPRI_MASK);
+		v = apic_read(APIC_PROCPRI);
+		printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
+	}
 
 	v = apic_read(APIC_EOI);
 	printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
@@ -1514,8 +1785,13 @@ __apicdebuginit(void) print_local_APIC(v
 	printk(KERN_DEBUG "... APIC IRR field:\n");
 	print_APIC_bitfield(APIC_IRR);
 
-	v = apic_read(APIC_ESR);
-	printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
+	if (APIC_INTEGRATED(ver)) {             /* !82489DX */
+		if (maxlvt > 3)         /* Due to the Pentium erratum 3AP. */
+			apic_write(APIC_ESR, 0);
+
+		v = apic_read(APIC_ESR);
+		printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
+	}
 
 	icr = apic_icr_read();
 	printk(KERN_DEBUG "... APIC ICR: %08x\n", icr);
@@ -1606,6 +1882,13 @@ void __init enable_IO_APIC(void)
 	int apic;
 	unsigned long flags;
 
+#ifdef CONFIG_X86_32
+	int i;
+	if (!pirqs_enabled)
+		for (i = 0; i < MAX_PIRQS; i++)
+			pirq_entries[i] = -1;
+#endif
+
 	/*
 	 * The number of IO-APIC IRQ registers (== #pins):
 	 */
@@ -1634,6 +1917,10 @@ void __init enable_IO_APIC(void)
 	}
  found_i8259:
 	/* Look to see what if the MP table has reported the ExtINT */
+	/* If we could not find the appropriate pin by looking at the ioapic
+	 * the i8259 probably is not connected the ioapic but give the
+	 * mptable a chance anyway.
+	 */
 	i8259_pin  = find_isa_irq_pin(0, mp_ExtINT);
 	i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
 	/* Trust the MP table if nothing is setup in the hardware */
@@ -1693,6 +1980,122 @@ void disable_IO_APIC(void)
 	disconnect_bsp_APIC(ioapic_i8259.pin != -1);
 }
 
+#ifdef CONFIG_X86_32
+/*
+ * function to set the IO-APIC physical IDs based on the
+ * values stored in the MPC table.
+ *
+ * by Matt Domsch <Matt_Domsch@dell.com>  Tue Dec 21 12:25:05 CST 1999
+ */
+
+static void __init setup_ioapic_ids_from_mpc(void)
+{
+	union IO_APIC_reg_00 reg_00;
+	physid_mask_t phys_id_present_map;
+	int apic;
+	int i;
+	unsigned char old_id;
+	unsigned long flags;
+
+	if (x86_quirks->setup_ioapic_ids && x86_quirks->setup_ioapic_ids())
+		return;
+
+	/*
+	 * Don't check I/O APIC IDs for xAPIC systems.  They have
+	 * no meaning without the serial APIC bus.
+	 */
+	if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
+		|| APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
+		return;
+	/*
+	 * This is broken; anything with a real cpu count has to
+	 * circumvent this idiocy regardless.
+	 */
+	phys_id_present_map = ioapic_phys_id_map(phys_cpu_present_map);
+
+	/*
+	 * 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].mp_apicid;
+
+		if (mp_ioapics[apic].mp_apicid >= get_physical_broadcast()) {
+			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n",
+				apic, mp_ioapics[apic].mp_apicid);
+			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
+				reg_00.bits.ID);
+			mp_ioapics[apic].mp_apicid = reg_00.bits.ID;
+		}
+
+		/*
+		 * Sanity check, is the ID really free? Every APIC in a
+		 * system must have a unique ID or we get lots of nice
+		 * 'stuck on smp_invalidate_needed IPI wait' messages.
+		 */
+		if (check_apicid_used(phys_id_present_map,
+					mp_ioapics[apic].mp_apicid)) {
+			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n",
+				apic, mp_ioapics[apic].mp_apicid);
+			for (i = 0; i < get_physical_broadcast(); i++)
+				if (!physid_isset(i, phys_id_present_map))
+					break;
+			if (i >= get_physical_broadcast())
+				panic("Max APIC ID exceeded!\n");
+			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
+				i);
+			physid_set(i, phys_id_present_map);
+			mp_ioapics[apic].mp_apicid = i;
+		} else {
+			physid_mask_t tmp;
+			tmp = apicid_to_cpu_present(mp_ioapics[apic].mp_apicid);
+			apic_printk(APIC_VERBOSE, "Setting %d in the "
+					"phys_id_present_map\n",
+					mp_ioapics[apic].mp_apicid);
+			physids_or(phys_id_present_map, phys_id_present_map, tmp);
+		}
+
+
+		/*
+		 * We need to adjust the IRQ routing table
+		 * if the ID changed.
+		 */
+		if (old_id != mp_ioapics[apic].mp_apicid)
+			for (i = 0; i < mp_irq_entries; i++)
+				if (mp_irqs[i].mp_dstapic == old_id)
+					mp_irqs[i].mp_dstapic
+						= mp_ioapics[apic].mp_apicid;
+
+		/*
+		 * 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].mp_apicid);
+
+		reg_00.bits.ID = mp_ioapics[apic].mp_apicid;
+		spin_lock_irqsave(&ioapic_lock, flags);
+
+		/*
+		 * Sanity check
+		 */
+		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].mp_apicid)
+			printk("could not set ID!\n");
+		else
+			apic_printk(APIC_VERBOSE, " ok.\n");
+	}
+}
+#endif
+
 int no_timer_check __initdata;
 
 static int __init notimercheck(char *s)
@@ -1778,8 +2181,10 @@ static unsigned int startup_ioapic_irq(u
 	return was_pending;
 }
 
+#ifdef CONFIG_X86_64
 static int ioapic_retrigger_irq(unsigned int irq)
 {
+
 	struct irq_cfg *cfg = irq_cfg(irq);
 	unsigned long flags;
 
@@ -1789,6 +2194,14 @@ static int ioapic_retrigger_irq(unsigned
 
 	return 1;
 }
+#else
+static int ioapic_retrigger_irq(unsigned int irq)
+{
+        send_IPI_self(irq_cfg(irq)->vector);
+
+        return 1;
+}
+#endif
 
 /*
  * Level and edge triggered IO-APIC interrupts need different handling,
@@ -1948,7 +2361,9 @@ asmlinkage void smp_irq_move_cleanup_int
 {
 	unsigned vector, me;
 	ack_APIC_irq();
+#ifdef CONFIG_X86_64
 	exit_idle();
+#endif
 	irq_enter();
 
 	me = smp_processor_id();
@@ -2020,6 +2435,7 @@ static void ack_apic_edge(unsigned int i
 	ack_APIC_irq();
 }
 
+#ifdef CONFIG_X86_64
 static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
 {
 	int do_unmask_irq = 0;
@@ -2072,6 +2488,49 @@ static void ack_apic_level(unsigned int
 		unmask_IO_APIC_irq(irq);
 	}
 }
+#else
+atomic_t irq_mis_count;
+static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
+{
+	unsigned long v;
+	int i;
+
+	irq_complete_move(irq);
+	move_native_irq(irq, desc);
+	/*
+	* 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
+	*/
+	i = irq_cfg(irq)->vector;
+
+	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
+
+	ack_APIC_irq();
+
+	if (!(v & (1 << (i & 0x1f)))) {
+		atomic_inc(&irq_mis_count);
+		spin_lock(&ioapic_lock);
+		__mask_and_edge_IO_APIC_irq(irq);
+		__unmask_and_level_IO_APIC_irq(irq);
+		spin_unlock(&ioapic_lock);
+	}
+}
+#endif
 
 static struct irq_chip ioapic_chip __read_mostly = {
 	.name 		= "IO-APIC",
@@ -2137,20 +2596,24 @@ static inline void init_IO_APIC_traps(vo
 	}
 }
 
-static void unmask_lapic_irq(unsigned int irq)
+/*
+ * The local APIC irq-chip implementation:
+ */
+
+static void mask_lapic_irq(unsigned int irq)
 {
 	unsigned long v;
 
 	v = apic_read(APIC_LVT0);
-	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
+	apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
 }
 
-static void mask_lapic_irq(unsigned int irq)
+static void unmask_lapic_irq(unsigned int irq)
 {
 	unsigned long v;
 
 	v = apic_read(APIC_LVT0);
-	apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
+	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
 }
 
 static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
@@ -2178,19 +2641,19 @@ static void lapic_register_intr(int irq)
 static void __init setup_nmi(void)
 {
 	/*
- 	 * Dirty trick to enable the NMI watchdog ...
+	 * Dirty trick to enable the NMI watchdog ...
 	 * We put the 8259A master into AEOI mode and
 	 * unmask on all local APICs LVT0 as NMI.
 	 *
 	 * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
 	 * is from Maciej W. Rozycki - so we do not have to EOI from
 	 * the NMI handler or the timer interrupt.
-	 */ 
-	printk(KERN_INFO "activating NMI Watchdog ...");
+	 */
+	apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ...");
 
 	enable_NMI_through_LVT0();
 
-	printk(" done.\n");
+	apic_printk(APIC_VERBOSE, " done.\n");
 }
 
 /*
@@ -2207,12 +2670,17 @@ static inline void __init unlock_ExtINT_
 	unsigned char save_control, save_freq_select;
 
 	pin  = find_isa_irq_pin(8, mp_INT);
+	if (pin == -1) {
+		WARN_ON_ONCE(1);
+		return;
+	}
 	apic = find_isa_irq_apic(8, mp_INT);
-	if (pin == -1)
+	if (apic == -1) {
+		WARN_ON_ONCE(1);
 		return;
+	}
 
 	entry0 = ioapic_read_entry(apic, pin);
-
 	clear_IO_APIC_pin(apic, pin);
 
 	memset(&entry1, 0, sizeof(entry1));
@@ -2264,17 +2732,21 @@ int timer_through_8259 __initdata;
  * is so screwy.  Thanks to Brian Perkins for testing/hacking this beast
  * fanatically on his truly buggy board.
  *
- * FIXME: really need to revamp this for modern platforms only.
+ * FIXME: really need to revamp this for all platforms.
  */
 static inline void __init check_timer(void)
 {
 	struct irq_cfg *cfg = irq_cfg(0);
 	int apic1, pin1, apic2, pin2;
 	unsigned long flags;
+	unsigned int ver;
 	int no_pin1 = 0;
 
 	local_irq_save(flags);
 
+        ver = apic_read(APIC_LVR);
+        ver = GET_APIC_VERSION(ver);
+
 	/*
 	 * get/set the timer IRQ vector:
 	 */
@@ -2283,10 +2755,18 @@ static inline void __init check_timer(vo
 
 	/*
 	 * As IRQ0 is to be enabled in the 8259A, the virtual
-	 * wire has to be disabled in the local APIC.
+	 * wire has to be disabled in the local APIC.  Also
+	 * timer interrupts need to be acknowledged manually in
+	 * the 8259A for the i82489DX when using the NMI
+	 * watchdog as that APIC treats NMIs as level-triggered.
+	 * The AEOI mode will finish them in the 8259A
+	 * automatically.
 	 */
 	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
 	init_8259A(1);
+#ifdef CONFIG_X86_32
+	timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver));
+#endif
 
 	pin1  = find_isa_irq_pin(0, mp_INT);
 	apic1 = find_isa_irq_apic(0, mp_INT);
@@ -2378,6 +2858,9 @@ static inline void __init check_timer(vo
 			    "through the IO-APIC - disabling NMI Watchdog!\n");
 		nmi_watchdog = NMI_NONE;
 	}
+#ifdef CONFIG_X86_32
+	timer_ack = 0;
+#endif
 
 	apic_printk(APIC_QUIET, KERN_INFO
 		    "...trying to set up timer as Virtual Wire IRQ...\n");
@@ -2431,19 +2914,29 @@ out:
  * the I/O APIC in all cases now.  No actual device should request
  * it anyway.  --macro
  */
-#define PIC_IRQS	(1<<2)
+#define PIC_IRQS	(1 << PIC_CASCADE_IR)
 
 void __init setup_IO_APIC(void)
 {
 
+#ifdef CONFIG_X86_32
+	enable_IO_APIC();
+#else
 	/*
 	 * calling enable_IO_APIC() is moved to setup_local_APIC for BP
 	 */
+#endif
 
 	io_apic_irqs = ~PIC_IRQS;
 
 	apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
-
+        /*
+         * Set up IO-APIC IRQ routing.
+         */
+#ifdef CONFIG_X86_32
+        if (!acpi_ioapic)
+                setup_ioapic_ids_from_mpc();
+#endif
 	sync_Arb_IDs();
 	setup_IO_APIC_irqs();
 	init_IO_APIC_traps();
@@ -3104,7 +3597,93 @@ int arch_setup_ht_irq(unsigned int irq,
 
 #ifdef CONFIG_ACPI
 
-#define IO_APIC_MAX_ID		0xFE
+#ifdef CONFIG_X86_32
+int __init io_apic_get_unique_id(int ioapic, int apic_id)
+{
+	union IO_APIC_reg_00 reg_00;
+	static physid_mask_t apic_id_map = PHYSID_MASK_NONE;
+	physid_mask_t tmp;
+	unsigned long flags;
+	int i = 0;
+
+	/*
+	 * The P4 platform supports up to 256 APIC IDs on two separate APIC
+	 * buses (one for LAPICs, one for IOAPICs), where predecessors only
+	 * supports up to 16 on one shared APIC bus.
+	 *
+	 * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full
+	 *      advantage of new APIC bus architecture.
+	 */
+
+	if (physids_empty(apic_id_map))
+		apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	reg_00.raw = io_apic_read(ioapic, 0);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	if (apic_id >= get_physical_broadcast()) {
+		printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying "
+			"%d\n", ioapic, apic_id, reg_00.bits.ID);
+		apic_id = reg_00.bits.ID;
+	}
+
+	/*
+	 * Every APIC in a system must have a unique ID or we get lots of nice
+	 * 'stuck on smp_invalidate_needed IPI wait' messages.
+	 */
+	if (check_apicid_used(apic_id_map, apic_id)) {
+
+		for (i = 0; i < get_physical_broadcast(); i++) {
+			if (!check_apicid_used(apic_id_map, i))
+				break;
+		}
+
+		if (i == get_physical_broadcast())
+			panic("Max apic_id exceeded!\n");
+
+		printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
+			"trying %d\n", ioapic, apic_id, i);
+
+		apic_id = i;
+	}
+
+	tmp = apicid_to_cpu_present(apic_id);
+	physids_or(apic_id_map, apic_id_map, tmp);
+
+	if (reg_00.bits.ID != apic_id) {
+		reg_00.bits.ID = apic_id;
+
+		spin_lock_irqsave(&ioapic_lock, flags);
+		io_apic_write(ioapic, 0, reg_00.raw);
+		reg_00.raw = io_apic_read(ioapic, 0);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+
+		/* Sanity check */
+		if (reg_00.bits.ID != apic_id) {
+			printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
+			return -1;
+		}
+	}
+
+	apic_printk(APIC_VERBOSE, KERN_INFO
+			"IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
+
+	return apic_id;
+}
+
+int __init io_apic_get_version(int ioapic)
+{
+	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);
+
+	return reg_01.bits.version;
+}
+#endif
 
 int __init io_apic_get_redir_entries (int ioapic)
 {
@@ -3206,6 +3785,7 @@ void __init setup_ioapic_dest(void)
 }
 #endif
 
+#ifdef CONFIG_X86_64
 #define IOAPIC_RESOURCE_NAME_SIZE 11
 
 static struct resource *ioapic_resources;
@@ -3241,36 +3821,56 @@ static struct resource * __init ioapic_s
 
 	return res;
 }
+#endif
 
 void __init ioapic_init_mappings(void)
 {
 	unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
-	struct resource *ioapic_res;
 	int i;
+#ifdef CONFIG_X86_64
+	struct resource *ioapic_res;
 
 	ioapic_res = ioapic_setup_resources();
+#endif
 	for (i = 0; i < nr_ioapics; i++) {
 		if (smp_found_config) {
 			ioapic_phys = mp_ioapics[i].mp_apicaddr;
+#ifdef CONFIG_X86_32
+                        if (!ioapic_phys) {
+                                printk(KERN_ERR
+                                       "WARNING: bogus zero IO-APIC "
+                                       "address found in MPTABLE, "
+                                       "disabling IO/APIC support!\n");
+                                smp_found_config = 0;
+                                skip_ioapic_setup = 1;
+                                goto fake_ioapic_page;
+                        }
+#endif
 		} else {
+#ifdef CONFIG_X86_32
+fake_ioapic_page:
+#endif
 			ioapic_phys = (unsigned long)
 				alloc_bootmem_pages(PAGE_SIZE);
 			ioapic_phys = __pa(ioapic_phys);
 		}
 		set_fixmap_nocache(idx, ioapic_phys);
 		apic_printk(APIC_VERBOSE,
-			    "mapped IOAPIC to %016lx (%016lx)\n",
+			    "mapped IOAPIC to %08lx (%08lx)\n",
 			    __fix_to_virt(idx), ioapic_phys);
 		idx++;
 
+#ifdef CONFIG_X86_64
 		if (ioapic_res != NULL) {
 			ioapic_res->start = ioapic_phys;
 			ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
 			ioapic_res++;
 		}
+#endif
 	}
 }
 
+#ifdef CONFIG_X86_64
 static int __init ioapic_insert_resources(void)
 {
 	int i;
@@ -3293,4 +3893,4 @@ static int __init ioapic_insert_resource
 /* Insert the IO APIC resources after PCI initialization has occured to handle
  * IO APICS that are mapped in on a BAR in PCI space. */
 late_initcall(ioapic_insert_resources);
-
+#endif

[PATCH 6/7] rename io_apic_64.c to io_apic.c

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/Makefile     |    2 
 arch/x86/kernel/io_apic.c    | 3898 +++++++++++++++++++++++++++++++++++++++++++
 arch/x86/kernel/io_apic_32.c | 3220 -----------------------------------
 arch/x86/kernel/io_apic_64.c | 3898 -------------------------------------------
 4 files changed, 3899 insertions(+), 7119 deletions(-)

Index: linux-2.6/arch/x86/kernel/Makefile
===================================================================
--- linux-2.6.orig/arch/x86/kernel/Makefile
+++ linux-2.6/arch/x86/kernel/Makefile
@@ -65,7 +65,7 @@ obj-$(CONFIG_X86_64_SMP)	+= tsc_sync.o s
 obj-$(CONFIG_X86_TRAMPOLINE)	+= trampoline_$(BITS).o
 obj-$(CONFIG_X86_MPPARSE)	+= mpparse.o
 obj-$(CONFIG_X86_LOCAL_APIC)	+= apic_$(BITS).o nmi.o
-obj-$(CONFIG_X86_IO_APIC)	+= io_apic_$(BITS).o
+obj-$(CONFIG_X86_IO_APIC)	+= io_apic.o
 obj-$(CONFIG_X86_REBOOTFIXUPS)	+= reboot_fixups_32.o
 obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
 obj-$(CONFIG_KEXEC)		+= machine_kexec_$(BITS).o
Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- /dev/null
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -0,0 +1,3896 @@
+/*
+ *	Intel IO-APIC support for multi-Pentium hosts.
+ *
+ *	Copyright (C) 1997, 1998, 1999, 2000 Ingo Molnar, Hajnalka Szabo
+ *
+ *	Many thanks to Stig Venaas for trying out countless experimental
+ *	patches and reporting/debugging problems patiently!
+ *
+ *	(c) 1999, Multiple IO-APIC support, developed by
+ *	Ken-ichi Yaku <yaku@css1.kbnes.nec.co.jp> and
+ *      Hidemi Kishimoto <kisimoto@css1.kbnes.nec.co.jp>,
+ *	further tested and cleaned up by Zach Brown <zab@redhat.com>
+ *	and Ingo Molnar <mingo@redhat.com>
+ *
+ *	Fixes
+ *	Maciej W. Rozycki	:	Bits for genuine 82489DX APICs;
+ *					thanks to Eric Gilmore
+ *					and Rolf G. Tews
+ *					for testing these extensively
+ *	Paul Diefenbaugh	:	Added full ACPI support
+ */
+
+#include <linux/mm.h>
+#include <linux/interrupt.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/sched.h>
+#include <linux/pci.h>
+#include <linux/mc146818rtc.h>
+#include <linux/compiler.h>
+#include <linux/acpi.h>
+#include <linux/module.h>
+#include <linux/sysdev.h>
+#include <linux/msi.h>
+#include <linux/htirq.h>
+#include <linux/freezer.h>
+#include <linux/kthread.h>
+#include <linux/jiffies.h>	/* time_after() */
+#ifdef CONFIG_ACPI
+#include <acpi/acpi_bus.h>
+#endif
+#include <linux/bootmem.h>
+#include <linux/dmar.h>
+
+#include <asm/idle.h>
+#include <asm/io.h>
+#include <asm/smp.h>
+#include <asm/desc.h>
+#include <asm/proto.h>
+#include <asm/acpi.h>
+#include <asm/dma.h>
+#include <asm/timer.h>
+#include <asm/i8259.h>
+#include <asm/nmi.h>
+#include <asm/msidef.h>
+#include <asm/hypertransport.h>
+#include <asm/setup.h>
+#include <asm/irq_remapping.h>
+
+#include <mach_ipi.h>
+#include <mach_apic.h>
+#include <mach_apicdef.h>
+
+#define __apicdebuginit(type) static type __init
+
+/*
+ *      Is the SiS APIC rmw bug present ?
+ *      -1 = don't know, 0 = no, 1 = yes
+ */
+int sis_apic_bug = -1;
+
+static DEFINE_SPINLOCK(ioapic_lock);
+static DEFINE_SPINLOCK(vector_lock);
+
+int first_free_entry;
+/*
+ * Rough estimation of how many shared IRQs there are, can
+ * be changed anytime.
+ */
+int pin_map_size;
+
+/*
+ * # of IRQ routing registers
+ */
+int nr_ioapic_registers[MAX_IO_APICS];
+
+/* I/O APIC entries */
+struct mp_config_ioapic mp_ioapics[MAX_IO_APICS];
+int nr_ioapics;
+
+/* MP IRQ source entries */
+struct mp_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
+
+/* # of MP IRQ source entries */
+int mp_irq_entries;
+
+#if defined (CONFIG_MCA) || defined (CONFIG_EISA)
+int mp_bus_id_to_type[MAX_MP_BUSSES];
+#endif
+
+DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES);
+
+int skip_ioapic_setup;
+
+static int __init parse_noapic(char *str)
+{
+	/* disable IO-APIC */
+	disable_ioapic_setup();
+	return 0;
+}
+early_param("noapic", parse_noapic);
+
+struct irq_cfg;
+struct irq_pin_list;
+struct irq_cfg {
+	unsigned int irq;
+	struct irq_cfg *next;
+	struct irq_pin_list *irq_2_pin;
+	cpumask_t domain;
+	cpumask_t old_domain;
+	unsigned move_cleanup_count;
+	u8 vector;
+	u8 move_in_progress : 1;
+};
+
+/* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
+static struct irq_cfg irq_cfg_legacy[] __initdata = {
+	[0]  = { .irq =  0, .domain = CPU_MASK_ALL, .vector = IRQ0_VECTOR,  },
+	[1]  = { .irq =  1, .domain = CPU_MASK_ALL, .vector = IRQ1_VECTOR,  },
+	[2]  = { .irq =  2, .domain = CPU_MASK_ALL, .vector = IRQ2_VECTOR,  },
+	[3]  = { .irq =  3, .domain = CPU_MASK_ALL, .vector = IRQ3_VECTOR,  },
+	[4]  = { .irq =  4, .domain = CPU_MASK_ALL, .vector = IRQ4_VECTOR,  },
+	[5]  = { .irq =  5, .domain = CPU_MASK_ALL, .vector = IRQ5_VECTOR,  },
+	[6]  = { .irq =  6, .domain = CPU_MASK_ALL, .vector = IRQ6_VECTOR,  },
+	[7]  = { .irq =  7, .domain = CPU_MASK_ALL, .vector = IRQ7_VECTOR,  },
+	[8]  = { .irq =  8, .domain = CPU_MASK_ALL, .vector = IRQ8_VECTOR,  },
+	[9]  = { .irq =  9, .domain = CPU_MASK_ALL, .vector = IRQ9_VECTOR,  },
+	[10] = { .irq = 10, .domain = CPU_MASK_ALL, .vector = IRQ10_VECTOR, },
+	[11] = { .irq = 11, .domain = CPU_MASK_ALL, .vector = IRQ11_VECTOR, },
+	[12] = { .irq = 12, .domain = CPU_MASK_ALL, .vector = IRQ12_VECTOR, },
+	[13] = { .irq = 13, .domain = CPU_MASK_ALL, .vector = IRQ13_VECTOR, },
+	[14] = { .irq = 14, .domain = CPU_MASK_ALL, .vector = IRQ14_VECTOR, },
+	[15] = { .irq = 15, .domain = CPU_MASK_ALL, .vector = IRQ15_VECTOR, },
+};
+
+static struct irq_cfg irq_cfg_init = { .irq =  -1U, };
+/* need to be biger than size of irq_cfg_legacy */
+static int nr_irq_cfg = 32;
+
+static int __init parse_nr_irq_cfg(char *arg)
+{
+	if (arg) {
+		nr_irq_cfg = simple_strtoul(arg, NULL, 0);
+		if (nr_irq_cfg < 32)
+			nr_irq_cfg = 32;
+	}
+	return 0;
+}
+
+early_param("nr_irq_cfg", parse_nr_irq_cfg);
+
+static void init_one_irq_cfg(struct irq_cfg *cfg)
+{
+	memcpy(cfg, &irq_cfg_init, sizeof(struct irq_cfg));
+}
+
+static struct irq_cfg *irq_cfgx;
+static struct irq_cfg *irq_cfgx_free;
+static void __init init_work(void *data)
+{
+	struct dyn_array *da = data;
+	struct irq_cfg *cfg;
+	int legacy_count;
+	int i;
+
+	cfg = *da->name;
+
+	memcpy(cfg, irq_cfg_legacy, sizeof(irq_cfg_legacy));
+
+	legacy_count = sizeof(irq_cfg_legacy)/sizeof(irq_cfg_legacy[0]);
+	for (i = legacy_count; i < *da->nr; i++)
+		init_one_irq_cfg(&cfg[i]);
+
+	for (i = 1; i < *da->nr; i++)
+		cfg[i-1].next = &cfg[i];
+
+	irq_cfgx_free = &irq_cfgx[legacy_count];
+	irq_cfgx[legacy_count - 1].next = NULL;
+}
+
+#define for_each_irq_cfg(cfg)		\
+	for (cfg = irq_cfgx; cfg; cfg = cfg->next)
+
+DEFINE_DYN_ARRAY(irq_cfgx, sizeof(struct irq_cfg), nr_irq_cfg, PAGE_SIZE, init_work);
+
+static struct irq_cfg *irq_cfg(unsigned int irq)
+{
+	struct irq_cfg *cfg;
+
+	cfg = irq_cfgx;
+	while (cfg) {
+		if (cfg->irq == irq)
+			return cfg;
+
+		cfg = cfg->next;
+	}
+
+	return NULL;
+}
+
+static struct irq_cfg *irq_cfg_with_new(unsigned int irq)
+{
+	struct irq_cfg *cfg, *cfg_pri;
+	int i;
+	int count = 0;
+
+	cfg_pri = cfg = irq_cfgx;
+	while (cfg) {
+		if (cfg->irq == irq)
+			return cfg;
+
+		cfg_pri = cfg;
+		cfg = cfg->next;
+		count++;
+	}
+
+	if (!irq_cfgx_free) {
+		unsigned long phys;
+		unsigned long total_bytes;
+		/*
+		 *  we run out of pre-allocate ones, allocate more
+		 */
+		printk(KERN_DEBUG "try to get more irq_cfg %d\n", nr_irq_cfg);
+
+		total_bytes = sizeof(struct irq_cfg) * nr_irq_cfg;
+		if (after_bootmem)
+			cfg = kzalloc(total_bytes, GFP_ATOMIC);
+		else
+			cfg = __alloc_bootmem_nopanic(total_bytes, PAGE_SIZE, 0);
+
+		if (!cfg)
+			panic("please boot with nr_irq_cfg= %d\n", count * 2);
+
+		phys = __pa(cfg);
+		printk(KERN_DEBUG "irq_irq ==> [%#lx - %#lx]\n", phys, phys + total_bytes);
+
+		for (i = 0; i < nr_irq_cfg; i++)
+			init_one_irq_cfg(&cfg[i]);
+
+		for (i = 1; i < nr_irq_cfg; i++)
+			cfg[i-1].next = &cfg[i];
+
+		irq_cfgx_free = cfg;
+	}
+
+	cfg = irq_cfgx_free;
+	irq_cfgx_free = irq_cfgx_free->next;
+	cfg->next = NULL;
+	if (cfg_pri)
+		cfg_pri->next = cfg;
+	else
+		irq_cfgx = cfg;
+	cfg->irq = irq;
+	printk(KERN_DEBUG "found new irq_cfg for irq %d\n", cfg->irq);
+#ifdef CONFIG_HAVE_SPARSE_IRQ_DEBUG
+	{
+		/* dump the results */
+		struct irq_cfg *cfg;
+		unsigned long phys;
+		unsigned long bytes = sizeof(struct irq_cfg);
+
+		printk(KERN_DEBUG "=========================== %d\n", irq);
+		printk(KERN_DEBUG "irq_cfg dump after get that for %d\n", irq);
+		for_each_irq_cfg(cfg) {
+			phys = __pa(cfg);
+			printk(KERN_DEBUG "irq_cfg %d ==> [%#lx - %#lx]\n", cfg->irq, phys, phys + bytes);
+		}
+		printk(KERN_DEBUG "===========================\n");
+	}
+#endif
+	return cfg;
+}
+
+/*
+ * 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 *irq_2_pin_head;
+/* fill one page ? */
+static int nr_irq_2_pin = 0x100;
+static struct irq_pin_list *irq_2_pin_ptr;
+static void __init irq_2_pin_init_work(void *data)
+{
+	struct dyn_array *da = data;
+	struct irq_pin_list *pin;
+	int i;
+
+	pin = *da->name;
+
+	for (i = 1; i < *da->nr; i++)
+		pin[i-1].next = &pin[i];
+
+	irq_2_pin_ptr = &pin[0];
+}
+DEFINE_DYN_ARRAY(irq_2_pin_head, sizeof(struct irq_pin_list), nr_irq_2_pin, PAGE_SIZE, irq_2_pin_init_work);
+
+static struct irq_pin_list *get_one_free_irq_2_pin(void)
+{
+	struct irq_pin_list *pin;
+	int i;
+
+	pin = irq_2_pin_ptr;
+
+	if (pin) {
+		irq_2_pin_ptr = pin->next;
+		pin->next = NULL;
+		return pin;
+	}
+
+	/*
+	 *  we run out of pre-allocate ones, allocate more
+	 */
+	printk(KERN_DEBUG "try to get more irq_2_pin %d\n", nr_irq_2_pin);
+
+	if (after_bootmem)
+		pin = kzalloc(sizeof(struct irq_pin_list)*nr_irq_2_pin,
+				 GFP_ATOMIC);
+	else
+		pin = __alloc_bootmem_nopanic(sizeof(struct irq_pin_list) *
+				nr_irq_2_pin, PAGE_SIZE, 0);
+
+	if (!pin)
+		panic("can not get more irq_2_pin\n");
+
+	for (i = 1; i < nr_irq_2_pin; i++)
+		pin[i-1].next = &pin[i];
+
+	irq_2_pin_ptr = pin->next;
+	pin->next = NULL;
+
+	return pin;
+}
+
+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].mp_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.
+ *
+ * Older SiS APIC requires we rewrite the index register
+ */
+static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
+{
+	struct io_apic __iomem *io_apic = io_apic_base(apic);
+        if (sis_apic_bug)
+                writel(reg, &io_apic->index);
+	writel(value, &io_apic->data);
+}
+
+#ifdef CONFIG_X86_64
+static bool io_apic_level_ack_pending(unsigned int irq)
+{
+	struct irq_pin_list *entry;
+	unsigned long flags;
+	struct irq_cfg *cfg = irq_cfg(irq);
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	entry = cfg->irq_2_pin;
+	for (;;) {
+		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? */
+		if (reg & IO_APIC_REDIR_REMOTE_IRR) {
+			spin_unlock_irqrestore(&ioapic_lock, flags);
+			return true;
+		}
+		if (!entry->next)
+			break;
+		entry = entry->next;
+	}
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	return false;
+}
+#endif
+
+union entry_union {
+	struct { u32 w1, w2; };
+	struct IO_APIC_route_entry entry;
+};
+
+static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin)
+{
+	union entry_union eu;
+	unsigned long flags;
+	spin_lock_irqsave(&ioapic_lock, flags);
+	eu.w1 = io_apic_read(apic, 0x10 + 2 * pin);
+	eu.w2 = io_apic_read(apic, 0x11 + 2 * pin);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+	return eu.entry;
+}
+
+/*
+ * When we write a new IO APIC routing entry, we need to write the high
+ * word first! If the mask bit in the low word is clear, we will enable
+ * the interrupt, and we need to make sure the entry is fully populated
+ * before that happens.
+ */
+static void
+__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
+{
+	union entry_union eu;
+	eu.entry = e;
+	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
+	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
+}
+
+static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&ioapic_lock, flags);
+	__ioapic_write_entry(apic, pin, e);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
+/*
+ * When we mask an IO APIC routing entry, we need to write the low
+ * word first, in order to set the mask bit before we change the
+ * high bits!
+ */
+static void ioapic_mask_entry(int apic, int pin)
+{
+	unsigned long flags;
+	union entry_union eu = { .entry.mask = 1 };
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
+	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
+#ifdef CONFIG_SMP
+static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, u8 vector)
+{
+	int apic, pin;
+	struct irq_cfg *cfg;
+	struct irq_pin_list *entry;
+
+	cfg = irq_cfg(irq);
+	entry = cfg->irq_2_pin;
+	for (;;) {
+		unsigned int reg;
+
+		if (!entry)
+			break;
+
+		apic = entry->apic;
+		pin = entry->pin;
+#ifdef CONFIG_INTR_REMAP
+		/*
+		 * With interrupt-remapping, destination information comes
+		 * from interrupt-remapping table entry.
+		 */
+		if (!irq_remapped(irq))
+			io_apic_write(apic, 0x11 + pin*2, dest);
+#else
+		io_apic_write(apic, 0x11 + pin*2, dest);
+#endif
+		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, cpumask_t mask);
+
+static void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	unsigned long flags;
+	unsigned int dest;
+	cpumask_t tmp;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	cfg = irq_cfg(irq);
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+	/*
+	 * Only the high 8 bits are valid.
+	 */
+	dest = SET_APIC_LOGICAL_ID(dest);
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	__target_IO_APIC_irq(irq, dest, cfg->vector);
+	desc->affinity = mask;
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+#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(unsigned int irq, int apic, int pin)
+{
+	struct irq_cfg *cfg;
+	struct irq_pin_list *entry;
+
+	/* first time to refer irq_cfg, so with new */
+	cfg = irq_cfg_with_new(irq);
+	entry = cfg->irq_2_pin;
+	if (!entry) {
+		entry = get_one_free_irq_2_pin();
+		cfg->irq_2_pin = entry;
+		entry->apic = apic;
+		entry->pin = pin;
+		printk(KERN_DEBUG " 0 add_pin_to_irq: irq %d --> apic %d pin %d\n", irq, apic, pin);
+		return;
+	}
+
+	while (entry->next) {
+		/* not again, please */
+		if (entry->apic == apic && entry->pin == pin)
+			return;
+
+		entry = entry->next;
+	}
+
+	entry->next = get_one_free_irq_2_pin();
+	entry = entry->next;
+	entry->apic = apic;
+	entry->pin = pin;
+	printk(KERN_DEBUG " x add_pin_to_irq: irq %d --> apic %d pin %d\n", irq, apic, pin);
+}
+
+/*
+ * Reroute an IRQ to a different pin.
+ */
+static void __init replace_pin_at_irq(unsigned int irq,
+				      int oldapic, int oldpin,
+				      int newapic, int newpin)
+{
+	struct irq_cfg *cfg = irq_cfg(irq);
+	struct irq_pin_list *entry = cfg->irq_2_pin;
+	int replaced = 0;
+
+	while (entry) {
+		if (entry->apic == oldapic && entry->pin == oldpin) {
+			entry->apic = newapic;
+			entry->pin = newpin;
+			replaced = 1;
+			/* every one is different, right? */
+			break;
+		}
+		entry = entry->next;
+	}
+
+	/* why? call replace before add? */
+	if (!replaced)
+		add_pin_to_irq(irq, newapic, newpin);
+}
+
+#ifdef CONFIG_X86_64
+/*
+ * Synchronize the IO-APIC and the CPU by doing
+ * a dummy read from the IO-APIC
+ */
+static inline void io_apic_sync(unsigned int apic)
+{
+	struct io_apic __iomem *io_apic = io_apic_base(apic);
+	readl(&io_apic->data);
+}
+
+#define __DO_ACTION(R, ACTION, FINAL)					\
+									\
+{									\
+	int pin;							\
+	struct irq_cfg *cfg;						\
+	struct irq_pin_list *entry;					\
+									\
+	cfg = irq_cfg(irq);						\
+	entry = cfg->irq_2_pin;						\
+	for (;;) {							\
+		unsigned int reg;					\
+		if (!entry)						\
+			break;						\
+		pin = entry->pin;					\
+		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
+		reg ACTION;						\
+		io_apic_modify(entry->apic, 0x10 + R + pin*2, reg);	\
+		FINAL;							\
+		if (!entry->next)					\
+			break;						\
+		entry = entry->next;					\
+	}								\
+}
+
+#define DO_ACTION(name,R,ACTION, FINAL)					\
+									\
+	static void name##_IO_APIC_irq (unsigned int irq)		\
+	__DO_ACTION(R, ACTION, FINAL)
+
+/* mask = 1 */
+DO_ACTION(__mask,	0, |= IO_APIC_REDIR_MASKED, io_apic_sync(entry->apic))
+
+/* mask = 0 */
+DO_ACTION(__unmask,	0, &= ~IO_APIC_REDIR_MASKED, )
+
+#else
+
+static void __modify_IO_APIC_irq(unsigned int irq, unsigned long enable, unsigned long disable)
+{
+	struct irq_cfg *cfg;
+	struct irq_pin_list *entry;
+	unsigned int pin, reg;
+
+	cfg = irq_cfg(irq);
+	entry = cfg->irq_2_pin;
+	for (;;) {
+		if (!entry)
+			break;
+		pin = entry->pin;
+		reg = io_apic_read(entry->apic, 0x10 + pin*2);
+		reg &= ~disable;
+		reg |= enable;
+		io_apic_modify(entry->apic, 0x10 + pin*2, reg);
+		if (!entry->next)
+			break;
+		entry = entry->next;
+	}
+}
+
+/* mask = 1 */
+static void __mask_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED, 0);
+}
+
+/* mask = 0 */
+static void __unmask_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, 0, IO_APIC_REDIR_MASKED);
+}
+
+/* mask = 1, trigger = 0 */
+static void __mask_and_edge_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED,
+				IO_APIC_REDIR_LEVEL_TRIGGER);
+}
+
+/* mask = 0, trigger = 1 */
+static void __unmask_and_level_IO_APIC_irq(unsigned int irq)
+{
+	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_LEVEL_TRIGGER,
+				IO_APIC_REDIR_MASKED);
+}
+
+#endif
+
+static void mask_IO_APIC_irq (unsigned int irq)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	__mask_IO_APIC_irq(irq);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
+static void unmask_IO_APIC_irq (unsigned int irq)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	__unmask_IO_APIC_irq(irq);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
+static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
+{
+	struct IO_APIC_route_entry entry;
+
+	/* Check delivery_mode to be sure we're not clearing an SMI pin */
+	entry = ioapic_read_entry(apic, pin);
+	if (entry.delivery_mode == dest_SMI)
+		return;
+	/*
+	 * Disable it in the IO-APIC irq-routing table:
+	 */
+	ioapic_mask_entry(apic, pin);
+}
+
+static void clear_IO_APIC (void)
+{
+	int apic, pin;
+
+	for (apic = 0; apic < nr_ioapics; apic++)
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
+			clear_IO_APIC_pin(apic, pin);
+}
+
+#if !defined(CONFIG_SMP) && defined(CONFIG_X86_32)
+void send_IPI_self(int vector)
+{
+	unsigned int cfg;
+
+	/*
+	 * Wait for idle.
+	 */
+	apic_wait_icr_idle();
+	cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
+	/*
+	 * Send the IPI. The write to APIC_ICR fires this off.
+	 */
+	apic_write(APIC_ICR, cfg);
+}
+#endif /* !CONFIG_SMP && CONFIG_X86_32*/
+
+#ifdef CONFIG_X86_32
+/*
+ * 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;
+
+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, KERN_INFO
+			"PIRQ redirection, working around broken MP-BIOS.\n");
+	max = MAX_PIRQS;
+	if (ints[0] < MAX_PIRQS)
+		max = ints[0];
+
+	for (i = 0; i < max; i++) {
+		apic_printk(APIC_VERBOSE, KERN_DEBUG
+				"... PIRQ%d -> IRQ %d\n", i, ints[i+1]);
+		/*
+		 * PIRQs are mapped upside down, usually.
+		 */
+		pirq_entries[MAX_PIRQS-i-1] = ints[i+1];
+	}
+	return 1;
+}
+
+__setup("pirq=", ioapic_pirq_setup);
+#endif /* CONFIG_X86_32 */
+
+#ifdef CONFIG_INTR_REMAP
+/* I/O APIC RTE contents at the OS boot up */
+static struct IO_APIC_route_entry *early_ioapic_entries[MAX_IO_APICS];
+
+/*
+ * Saves and masks all the unmasked IO-APIC RTE's
+ */
+int save_mask_IO_APIC_setup(void)
+{
+	union IO_APIC_reg_01 reg_01;
+	unsigned long flags;
+	int apic, pin;
+
+	/*
+	 * The number of IO-APIC IRQ registers (== #pins):
+	 */
+	for (apic = 0; apic < nr_ioapics; apic++) {
+		spin_lock_irqsave(&ioapic_lock, flags);
+		reg_01.raw = io_apic_read(apic, 1);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
+	}
+
+	for (apic = 0; apic < nr_ioapics; apic++) {
+		early_ioapic_entries[apic] =
+			kzalloc(sizeof(struct IO_APIC_route_entry) *
+				nr_ioapic_registers[apic], GFP_KERNEL);
+		if (!early_ioapic_entries[apic])
+			return -ENOMEM;
+	}
+
+	for (apic = 0; apic < nr_ioapics; apic++)
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+			struct IO_APIC_route_entry entry;
+
+			entry = early_ioapic_entries[apic][pin] =
+				ioapic_read_entry(apic, pin);
+			if (!entry.mask) {
+				entry.mask = 1;
+				ioapic_write_entry(apic, pin, entry);
+			}
+		}
+	return 0;
+}
+
+void restore_IO_APIC_setup(void)
+{
+	int apic, pin;
+
+	for (apic = 0; apic < nr_ioapics; apic++)
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
+			ioapic_write_entry(apic, pin,
+					   early_ioapic_entries[apic][pin]);
+}
+
+void reinit_intr_remapped_IO_APIC(int intr_remapping)
+{
+	/*
+	 * 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();
+}
+#endif
+
+/*
+ * Find the IRQ entry number of a certain pin.
+ */
+static int find_irq_entry(int apic, int pin, int type)
+{
+	int i;
+
+	for (i = 0; i < mp_irq_entries; i++)
+		if (mp_irqs[i].mp_irqtype == type &&
+		    (mp_irqs[i].mp_dstapic == mp_ioapics[apic].mp_apicid ||
+		     mp_irqs[i].mp_dstapic == MP_APIC_ALL) &&
+		    mp_irqs[i].mp_dstirq == pin)
+			return i;
+
+	return -1;
+}
+
+/*
+ * Find the pin to which IRQ[irq] (ISA) is connected
+ */
+static int __init find_isa_irq_pin(int irq, int type)
+{
+	int i;
+
+	for (i = 0; i < mp_irq_entries; i++) {
+		int lbus = mp_irqs[i].mp_srcbus;
+
+		if (test_bit(lbus, mp_bus_not_pci) &&
+		    (mp_irqs[i].mp_irqtype == type) &&
+		    (mp_irqs[i].mp_srcbusirq == irq))
+
+			return mp_irqs[i].mp_dstirq;
+	}
+	return -1;
+}
+
+static int __init find_isa_irq_apic(int irq, int type)
+{
+	int i;
+
+	for (i = 0; i < mp_irq_entries; i++) {
+		int lbus = mp_irqs[i].mp_srcbus;
+
+		if (test_bit(lbus, mp_bus_not_pci) &&
+		    (mp_irqs[i].mp_irqtype == type) &&
+		    (mp_irqs[i].mp_srcbusirq == irq))
+			break;
+	}
+	if (i < mp_irq_entries) {
+		int apic;
+		for(apic = 0; apic < nr_ioapics; apic++) {
+			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic)
+				return apic;
+		}
+	}
+
+	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].mp_srcbus;
+
+		for (apic = 0; apic < nr_ioapics; apic++)
+			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic ||
+			    mp_irqs[i].mp_dstapic == MP_APIC_ALL)
+				break;
+
+		if (!test_bit(lbus, mp_bus_not_pci) &&
+		    !mp_irqs[i].mp_irqtype &&
+		    (bus == lbus) &&
+		    (slot == ((mp_irqs[i].mp_srcbusirq >> 2) & 0x1f))) {
+			int irq = pin_2_irq(i,apic,mp_irqs[i].mp_dstirq);
+
+			if (!(apic || IO_APIC_IRQ(irq)))
+				continue;
+
+			if (pin == (mp_irqs[i].mp_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
+ */
+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, KERN_INFO
+			"Broken MPtable reports ISA irq %d\n", irq);
+	return 0;
+}
+
+#endif
+
+/* ISA interrupts are always polarity zero edge triggered,
+ * when listed as conforming in the MP table. */
+
+#define default_ISA_trigger(idx)	(0)
+#define default_ISA_polarity(idx)	(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].mp_srcbusirq))
+#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
+
+/* PCI interrupts are always polarity one level triggered,
+ * when listed as conforming in the MP table. */
+
+#define default_PCI_trigger(idx)	(1)
+#define default_PCI_polarity(idx)	(1)
+
+/* 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].mp_srcbusirq))
+#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
+
+
+static int MPBIOS_polarity(int idx)
+{
+	int bus = mp_irqs[idx].mp_srcbus;
+	int polarity;
+
+	/*
+	 * Determine IRQ line polarity (high active or low active):
+	 */
+	switch (mp_irqs[idx].mp_irqflag & 3)
+	{
+		case 0: /* conforms, ie. bus-type dependent polarity */
+			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;
+			break;
+		}
+		case 2: /* reserved */
+		{
+			printk(KERN_WARNING "broken BIOS!!\n");
+			polarity = 1;
+			break;
+		}
+		case 3: /* low active */
+		{
+			polarity = 1;
+			break;
+		}
+		default: /* invalid */
+		{
+			printk(KERN_WARNING "broken BIOS!!\n");
+			polarity = 1;
+			break;
+		}
+	}
+	return polarity;
+}
+
+static int MPBIOS_trigger(int idx)
+{
+	int bus = mp_irqs[idx].mp_srcbus;
+	int trigger;
+
+	/*
+	 * Determine IRQ trigger mode (edge or level sensitive):
+	 */
+	switch ((mp_irqs[idx].mp_irqflag>>2) & 3)
+	{
+		case 0: /* conforms, ie. bus-type dependent */
+			if (test_bit(bus, mp_bus_not_pci))
+				trigger = default_ISA_trigger(idx);
+			else
+				trigger = default_PCI_trigger(idx);
+#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
+			switch (mp_bus_id_to_type[bus]) {
+				case MP_BUS_ISA: /* ISA pin */
+				{
+					/* set before the switch */
+					break;
+				}
+				case MP_BUS_EISA: /* EISA pin */
+				{
+					trigger = default_EISA_trigger(idx);
+					break;
+				}
+				case MP_BUS_PCI: /* PCI pin */
+				{
+					/* set before the switch */
+					break;
+				}
+				case MP_BUS_MCA: /* MCA pin */
+				{
+					trigger = default_MCA_trigger(idx);
+					break;
+				}
+				default:
+				{
+					printk(KERN_WARNING "broken BIOS!!\n");
+					trigger = 1;
+					break;
+				}
+			}
+#endif
+			break;
+		case 1: /* edge */
+		{
+			trigger = 0;
+			break;
+		}
+		case 2: /* reserved */
+		{
+			printk(KERN_WARNING "broken BIOS!!\n");
+			trigger = 1;
+			break;
+		}
+		case 3: /* level */
+		{
+			trigger = 1;
+			break;
+		}
+		default: /* invalid */
+		{
+			printk(KERN_WARNING "broken BIOS!!\n");
+			trigger = 0;
+			break;
+		}
+	}
+	return trigger;
+}
+
+static inline int irq_polarity(int idx)
+{
+	return MPBIOS_polarity(idx);
+}
+
+static inline int irq_trigger(int idx)
+{
+	return MPBIOS_trigger(idx);
+}
+
+int (*ioapic_renumber_irq)(int ioapic, int irq);
+static int pin_2_irq(int idx, int apic, int pin)
+{
+	int irq, i;
+	int bus = mp_irqs[idx].mp_srcbus;
+
+	/*
+	 * Debugging check, we are in big trouble if this message pops up!
+	 */
+	if (mp_irqs[idx].mp_dstirq != pin)
+		printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
+
+	if (test_bit(bus, mp_bus_not_pci)) {
+		irq = mp_irqs[idx].mp_srcbusirq;
+	} else {
+		/*
+		 * PCI IRQs are mapped in order
+		 */
+		i = irq = 0;
+		while (i < apic)
+			irq += nr_ioapic_registers[i++];
+		irq += pin;
+                /*
+                 * For MPS mode, so far only needed by ES7000 platform
+                 */
+                if (ioapic_renumber_irq)
+                        irq = ioapic_renumber_irq(apic, irq);
+	}
+
+#ifdef CONFIG_X86_32
+	/*
+	 * PCI IRQ command line redirection. Yes, limits are hardcoded.
+	 */
+	if ((pin >= 16) && (pin <= 23)) {
+		if (pirq_entries[pin-16] != -1) {
+			if (!pirq_entries[pin-16]) {
+				apic_printk(APIC_VERBOSE, KERN_DEBUG
+						"disabling PIRQ%d\n", pin-16);
+			} else {
+				irq = pirq_entries[pin-16];
+				apic_printk(APIC_VERBOSE, KERN_DEBUG
+						"using PIRQ%d -> IRQ %d\n",
+						pin-16, irq);
+			}
+		}
+	}
+#endif
+
+	return irq;
+}
+
+void lock_vector_lock(void)
+{
+	/* Used to the online set of cpus does not change
+	 * during assign_irq_vector.
+	 */
+	spin_lock(&vector_lock);
+}
+
+void unlock_vector_lock(void)
+{
+	spin_unlock(&vector_lock);
+}
+
+static int __assign_irq_vector(int irq, cpumask_t mask)
+{
+	/*
+	 * NOTE! The local APIC isn't very good at handling
+	 * multiple interrupts at the same interrupt level.
+	 * As the interrupt level is determined by taking the
+	 * vector number and shifting that right by 4, we
+	 * want to spread these out a bit so that they don't
+	 * all fall in the same interrupt level.
+	 *
+	 * Also, we've got to be careful not to trash gate
+	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
+	 */
+	static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0;
+	unsigned int old_vector;
+	int cpu;
+	struct irq_cfg *cfg;
+
+	cfg = irq_cfg(irq);
+
+	/* Only try and allocate irqs on cpus that are present */
+	cpus_and(mask, mask, cpu_online_map);
+
+	if ((cfg->move_in_progress) || cfg->move_cleanup_count)
+		return -EBUSY;
+
+	old_vector = cfg->vector;
+	if (old_vector) {
+		cpumask_t tmp;
+		cpus_and(tmp, cfg->domain, mask);
+		if (!cpus_empty(tmp))
+			return 0;
+	}
+
+	for_each_cpu_mask_nr(cpu, mask) {
+		cpumask_t domain, new_mask;
+		int new_cpu;
+		int vector, offset;
+
+		domain = vector_allocation_domain(cpu);
+		cpus_and(new_mask, domain, cpu_online_map);
+
+		vector = current_vector;
+		offset = current_offset;
+next:
+		vector += 8;
+		if (vector >= first_system_vector) {
+			/* If we run out of vectors on large boxen, must share them. */
+			offset = (offset + 1) % 8;
+			vector = FIRST_DEVICE_VECTOR + offset;
+		}
+		if (unlikely(current_vector == vector))
+			continue;
+#ifdef CONFIG_X86_64
+		if (vector == IA32_SYSCALL_VECTOR)
+			goto next;
+#else
+		if (vector == SYSCALL_VECTOR)
+			goto next;
+#endif
+		for_each_cpu_mask_nr(new_cpu, new_mask)
+			if (per_cpu(vector_irq, new_cpu)[vector] != -1)
+				goto next;
+		/* Found one! */
+		current_vector = vector;
+		current_offset = offset;
+		if (old_vector) {
+			cfg->move_in_progress = 1;
+			cfg->old_domain = cfg->domain;
+		}
+		for_each_cpu_mask_nr(new_cpu, new_mask)
+			per_cpu(vector_irq, new_cpu)[vector] = irq;
+		cfg->vector = vector;
+		cfg->domain = domain;
+		return 0;
+	}
+	return -ENOSPC;
+}
+
+static int assign_irq_vector(int irq, cpumask_t mask)
+{
+	int err;
+	unsigned long flags;
+
+	spin_lock_irqsave(&vector_lock, flags);
+	err = __assign_irq_vector(irq, mask);
+	spin_unlock_irqrestore(&vector_lock, flags);
+	return err;
+}
+
+static void __clear_irq_vector(int irq)
+{
+	struct irq_cfg *cfg;
+	cpumask_t mask;
+	int cpu, vector;
+
+	cfg = irq_cfg(irq);
+	BUG_ON(!cfg->vector);
+
+	vector = cfg->vector;
+	cpus_and(mask, cfg->domain, cpu_online_map);
+	for_each_cpu_mask_nr(cpu, mask)
+		per_cpu(vector_irq, cpu)[vector] = -1;
+
+	cfg->vector = 0;
+	cpus_clear(cfg->domain);
+}
+
+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;
+	struct irq_cfg *cfg;
+
+	/* Mark the inuse vectors */
+	for_each_irq_cfg(cfg) {
+		if (!cpu_isset(cpu, cfg->domain))
+			continue;
+		vector = cfg->vector;
+		irq = cfg->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;
+
+		cfg = irq_cfg(irq);
+		if (!cpu_isset(cpu, cfg->domain))
+			per_cpu(vector_irq, cpu)[vector] = -1;
+	}
+}
+
+static struct irq_chip ioapic_chip;
+#ifdef CONFIG_INTR_REMAP
+static struct irq_chip ir_ioapic_chip;
+#endif
+
+#define IOAPIC_AUTO     -1
+#define IOAPIC_EDGE     0
+#define IOAPIC_LEVEL    1
+
+#ifdef CONFIG_X86_32
+static inline int IO_APIC_irq_trigger(int irq)
+{
+        int apic, idx, pin;
+
+        for (apic = 0; apic < nr_ioapics; apic++) {
+                for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+                        idx = find_irq_entry(apic, pin, mp_INT);
+                        if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
+                                return irq_trigger(idx);
+                }
+        }
+        /*
+         * nonexistent IRQs are edge default
+         */
+        return 0;
+}
+#else
+static inline int IO_APIC_irq_trigger(int irq)
+{
+	return 1;
+}
+#endif
+
+static void ioapic_register_intr(int irq, unsigned long trigger)
+{
+	struct irq_desc *desc;
+
+	/* first time to use this irq_desc */
+	if (irq < 16)
+		desc = irq_desc(irq);
+	else
+		desc = irq_desc_with_new(irq);
+
+	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
+	    trigger == IOAPIC_LEVEL)
+		desc->status |= IRQ_LEVEL;
+	else
+		desc->status &= ~IRQ_LEVEL;
+
+#ifdef CONFIG_INTR_REMAP
+	if (irq_remapped(irq)) {
+		desc->status |= IRQ_MOVE_PCNTXT;
+		if (trigger)
+			set_irq_chip_and_handler_name(irq, &ir_ioapic_chip,
+						      handle_fasteoi_irq,
+						     "fasteoi");
+		else
+			set_irq_chip_and_handler_name(irq, &ir_ioapic_chip,
+						      handle_edge_irq, "edge");
+		return;
+	}
+#endif
+	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
+	    trigger == IOAPIC_LEVEL)
+		set_irq_chip_and_handler_name(irq, &ioapic_chip,
+					      handle_fasteoi_irq,
+					      "fasteoi");
+	else
+		set_irq_chip_and_handler_name(irq, &ioapic_chip,
+					      handle_edge_irq, "edge");
+}
+
+static int setup_ioapic_entry(int apic, int irq,
+			      struct IO_APIC_route_entry *entry,
+			      unsigned int destination, int trigger,
+			      int polarity, int vector)
+{
+	/*
+	 * add it to the IO-APIC irq-routing table:
+	 */
+	memset(entry,0,sizeof(*entry));
+
+#ifdef CONFIG_INTR_REMAP
+	if (intr_remapping_enabled) {
+		struct intel_iommu *iommu = map_ioapic_to_ir(apic);
+		struct irte irte;
+		struct IR_IO_APIC_route_entry *ir_entry =
+			(struct IR_IO_APIC_route_entry *) entry;
+		int index;
+
+		if (!iommu)
+			panic("No mapping iommu for ioapic %d\n", apic);
+
+		index = alloc_irte(iommu, irq, 1);
+		if (index < 0)
+			panic("Failed to allocate IRTE for ioapic %d\n", apic);
+
+		memset(&irte, 0, sizeof(irte));
+
+		irte.present = 1;
+		irte.dst_mode = INT_DEST_MODE;
+		irte.trigger_mode = trigger;
+		irte.dlvry_mode = INT_DELIVERY_MODE;
+		irte.vector = vector;
+		irte.dest_id = IRTE_DEST(destination);
+
+		modify_irte(irq, &irte);
+
+		ir_entry->index2 = (index >> 15) & 0x1;
+		ir_entry->zero = 0;
+		ir_entry->format = 1;
+		ir_entry->index = (index & 0x7fff);
+	} else
+#endif
+	{
+		entry->delivery_mode = INT_DELIVERY_MODE;
+		entry->dest_mode = INT_DEST_MODE;
+		entry->dest = destination;
+	}
+
+	entry->mask = 0;				/* enable IRQ */
+	entry->trigger = trigger;
+	entry->polarity = polarity;
+	entry->vector = vector;
+
+	/* Mask level triggered irqs.
+	 * Use IRQ_DELAYED_DISABLE for edge triggered irqs.
+	 */
+	if (trigger)
+		entry->mask = 1;
+	return 0;
+}
+
+static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq,
+			      int trigger, int polarity)
+{
+	struct irq_cfg *cfg;
+	struct IO_APIC_route_entry entry;
+	cpumask_t mask;
+
+	if (!IO_APIC_IRQ(irq))
+		return;
+
+	cfg = irq_cfg(irq);
+
+	mask = TARGET_CPUS;
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cpus_and(mask, cfg->domain, mask);
+
+	apic_printk(APIC_VERBOSE,KERN_DEBUG
+		    "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> "
+		    "IRQ %d Mode:%i Active:%i)\n",
+		    apic, mp_ioapics[apic].mp_apicid, pin, cfg->vector,
+		    irq, trigger, polarity);
+
+
+	if (setup_ioapic_entry(mp_ioapics[apic].mp_apicid, irq, &entry,
+			       cpu_mask_to_apicid(mask), trigger, polarity,
+			       cfg->vector)) {
+		printk("Failed to setup ioapic entry for ioapic  %d, pin %d\n",
+		       mp_ioapics[apic].mp_apicid, pin);
+		__clear_irq_vector(irq);
+		return;
+	}
+
+	ioapic_register_intr(irq, trigger);
+	if (irq < 16)
+		disable_8259A_irq(irq);
+
+	ioapic_write_entry(apic, pin, entry);
+}
+
+static void __init setup_IO_APIC_irqs(void)
+{
+	int apic, pin, idx, irq, first_notcon = 1;
+
+	apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
+
+	for (apic = 0; apic < nr_ioapics; apic++) {
+	for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+
+		idx = find_irq_entry(apic,pin,mp_INT);
+		if (idx == -1) {
+			if (first_notcon) {
+				apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mp_apicid, pin);
+				first_notcon = 0;
+			} else
+				apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mp_apicid, pin);
+			continue;
+		}
+		if (!first_notcon) {
+			apic_printk(APIC_VERBOSE, " not connected.\n");
+			first_notcon = 1;
+		}
+
+		irq = pin_2_irq(idx, apic, pin);
+#ifdef CONFIG_X86_32
+                if (multi_timer_check(apic, irq))
+                        continue;
+#endif
+		add_pin_to_irq(irq, apic, pin);
+
+		setup_IO_APIC_irq(apic, pin, irq,
+				  irq_trigger(idx), irq_polarity(idx));
+	}
+	}
+
+	if (!first_notcon)
+		apic_printk(APIC_VERBOSE, " not connected.\n");
+}
+
+/*
+ * Set up the timer pin, possibly with the 8259A-master behind.
+ */
+static void __init setup_timer_IRQ0_pin(unsigned int apic, unsigned int pin,
+					int vector)
+{
+	struct IO_APIC_route_entry entry;
+
+#ifdef CONFIG_INTR_REMAP
+	if (intr_remapping_enabled)
+		return;
+#endif
+
+	memset(&entry, 0, sizeof(entry));
+
+	/*
+	 * We use logical delivery to get the timer IRQ
+	 * to the first CPU.
+	 */
+	entry.dest_mode = INT_DEST_MODE;
+	entry.mask = 1;					/* mask IRQ now */
+	entry.dest = cpu_mask_to_apicid(TARGET_CPUS);
+	entry.delivery_mode = INT_DELIVERY_MODE;
+	entry.polarity = 0;
+	entry.trigger = 0;
+	entry.vector = vector;
+
+	/*
+	 * The timer IRQ doesn't have to know that behind the
+	 * scene we may have a 8259A-master in AEOI mode ...
+	 */
+	set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge");
+
+	/*
+	 * Add it to the IO-APIC irq-routing table:
+	 */
+	ioapic_write_entry(apic, pin, entry);
+}
+
+
+__apicdebuginit(void) print_IO_APIC(void)
+{
+	int apic, i;
+	union IO_APIC_reg_00 reg_00;
+	union IO_APIC_reg_01 reg_01;
+	union IO_APIC_reg_02 reg_02;
+	union IO_APIC_reg_03 reg_03;
+	unsigned long flags;
+	struct irq_cfg *cfg;
+
+	if (apic_verbosity == APIC_QUIET)
+		return;
+
+	printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries);
+	for (i = 0; i < nr_ioapics; i++)
+		printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n",
+		       mp_ioapics[i].mp_apicid, nr_ioapic_registers[i]);
+
+	/*
+	 * We are a bit conservative about what we expect.  We have to
+	 * know about every hardware change ASAP.
+	 */
+	printk(KERN_INFO "testing the IO APIC.......................\n");
+
+	for (apic = 0; apic < nr_ioapics; apic++) {
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	reg_00.raw = io_apic_read(apic, 0);
+	reg_01.raw = io_apic_read(apic, 1);
+	if (reg_01.bits.version >= 0x10)
+		reg_02.raw = io_apic_read(apic, 2);
+        if (reg_01.bits.version >= 0x20)
+                reg_03.raw = io_apic_read(apic, 3);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	printk("\n");
+	printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid);
+	printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
+	printk(KERN_DEBUG ".......    : physical APIC id: %02X\n", reg_00.bits.ID);
+	printk(KERN_DEBUG ".......    : Delivery Type: %X\n", reg_00.bits.delivery_type);
+	printk(KERN_DEBUG ".......    : LTS          : %X\n", reg_00.bits.LTS);
+
+	printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)&reg_01);
+	printk(KERN_DEBUG ".......     : max redirection entries: %04X\n", reg_01.bits.entries);
+
+	printk(KERN_DEBUG ".......     : PRQ implemented: %X\n", reg_01.bits.PRQ);
+	printk(KERN_DEBUG ".......     : IO APIC version: %04X\n", reg_01.bits.version);
+
+	/*
+	 * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02,
+	 * but the value of reg_02 is read as the previous read register
+	 * value, so ignore it if reg_02 == reg_01.
+	 */
+	if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) {
+		printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
+		printk(KERN_DEBUG ".......     : arbitration: %02X\n", reg_02.bits.arbitration);
+	}
+
+	/*
+	 * Some Intel chipsets with IO APIC VERSION of 0x2? don't have reg_02
+	 * or reg_03, but the value of reg_0[23] is read as the previous read
+	 * register value, so ignore it if reg_03 == reg_0[12].
+	 */
+	if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw &&
+	    reg_03.raw != reg_01.raw) {
+		printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw);
+		printk(KERN_DEBUG ".......     : Boot DT    : %X\n", reg_03.bits.boot_DT);
+	}
+
+	printk(KERN_DEBUG ".... IRQ redirection table:\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;
+
+		entry = ioapic_read_entry(apic, i);
+
+		printk(KERN_DEBUG " %02x %03X ",
+			i,
+			entry.dest
+		);
+
+		printk("%1d    %1d    %1d   %1d   %1d    %1d    %1d    %02X\n",
+			entry.mask,
+			entry.trigger,
+			entry.irr,
+			entry.polarity,
+			entry.delivery_status,
+			entry.dest_mode,
+			entry.delivery_mode,
+			entry.vector
+		);
+	}
+	}
+	printk(KERN_DEBUG "IRQ to pin mappings:\n");
+	for_each_irq_cfg(cfg) {
+		struct irq_pin_list *entry = cfg->irq_2_pin;
+		if (!entry)
+			continue;
+		printk(KERN_DEBUG "IRQ%d ", cfg->irq);
+		for (;;) {
+			printk("-> %d:%d", entry->apic, entry->pin);
+			if (!entry->next)
+				break;
+			entry = entry->next;
+		}
+		printk("\n");
+	}
+
+	printk(KERN_INFO ".................................... done.\n");
+
+	return;
+}
+
+__apicdebuginit(void) print_APIC_bitfield(int base)
+{
+	unsigned int v;
+	int i, j;
+
+	if (apic_verbosity == APIC_QUIET)
+		return;
+
+	printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
+	for (i = 0; i < 8; i++) {
+		v = apic_read(base + i*0x10);
+		for (j = 0; j < 32; j++) {
+			if (v & (1<<j))
+				printk("1");
+			else
+				printk("0");
+		}
+		printk("\n");
+	}
+}
+
+__apicdebuginit(void) print_local_APIC(void *dummy)
+{
+	unsigned int v, ver, maxlvt;
+	u64 icr;
+
+	if (apic_verbosity == APIC_QUIET)
+		return;
+
+	printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
+		smp_processor_id(), hard_smp_processor_id());
+	v = apic_read(APIC_ID);
+	printk(KERN_INFO "... APIC ID:      %08x (%01x)\n", v, read_apic_id());
+	v = apic_read(APIC_LVR);
+	printk(KERN_INFO "... APIC VERSION: %08x\n", v);
+	ver = GET_APIC_VERSION(v);
+	maxlvt = lapic_get_maxlvt();
+
+	v = apic_read(APIC_TASKPRI);
+	printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
+
+	if (APIC_INTEGRATED(ver)) {                     /* !82489DX */
+		v = apic_read(APIC_ARBPRI);
+		printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
+			v & APIC_ARBPRI_MASK);
+		v = apic_read(APIC_PROCPRI);
+		printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
+	}
+
+	v = apic_read(APIC_EOI);
+	printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
+	v = apic_read(APIC_RRR);
+	printk(KERN_DEBUG "... APIC RRR: %08x\n", v);
+	v = apic_read(APIC_LDR);
+	printk(KERN_DEBUG "... APIC LDR: %08x\n", v);
+	v = apic_read(APIC_DFR);
+	printk(KERN_DEBUG "... APIC DFR: %08x\n", v);
+	v = apic_read(APIC_SPIV);
+	printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
+
+	printk(KERN_DEBUG "... APIC ISR field:\n");
+	print_APIC_bitfield(APIC_ISR);
+	printk(KERN_DEBUG "... APIC TMR field:\n");
+	print_APIC_bitfield(APIC_TMR);
+	printk(KERN_DEBUG "... APIC IRR field:\n");
+	print_APIC_bitfield(APIC_IRR);
+
+	if (APIC_INTEGRATED(ver)) {             /* !82489DX */
+		if (maxlvt > 3)         /* Due to the Pentium erratum 3AP. */
+			apic_write(APIC_ESR, 0);
+
+		v = apic_read(APIC_ESR);
+		printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
+	}
+
+	icr = apic_icr_read();
+	printk(KERN_DEBUG "... APIC ICR: %08x\n", icr);
+	printk(KERN_DEBUG "... APIC ICR2: %08x\n", icr >> 32);
+
+	v = apic_read(APIC_LVTT);
+	printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
+
+	if (maxlvt > 3) {                       /* PC is LVT#4. */
+		v = apic_read(APIC_LVTPC);
+		printk(KERN_DEBUG "... APIC LVTPC: %08x\n", v);
+	}
+	v = apic_read(APIC_LVT0);
+	printk(KERN_DEBUG "... APIC LVT0: %08x\n", v);
+	v = apic_read(APIC_LVT1);
+	printk(KERN_DEBUG "... APIC LVT1: %08x\n", v);
+
+	if (maxlvt > 2) {			/* ERR is LVT#3. */
+		v = apic_read(APIC_LVTERR);
+		printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
+	}
+
+	v = apic_read(APIC_TMICT);
+	printk(KERN_DEBUG "... APIC TMICT: %08x\n", v);
+	v = apic_read(APIC_TMCCT);
+	printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
+	v = apic_read(APIC_TDCR);
+	printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
+	printk("\n");
+}
+
+__apicdebuginit(void) print_all_local_APICs(void)
+{
+	on_each_cpu(print_local_APIC, NULL, 1);
+}
+
+__apicdebuginit(void) print_PIC(void)
+{
+	unsigned int v;
+	unsigned long flags;
+
+	if (apic_verbosity == APIC_QUIET)
+		return;
+
+	printk(KERN_DEBUG "\nprinting PIC contents\n");
+
+	spin_lock_irqsave(&i8259A_lock, flags);
+
+	v = inb(0xa1) << 8 | inb(0x21);
+	printk(KERN_DEBUG "... PIC  IMR: %04x\n", v);
+
+	v = inb(0xa0) << 8 | inb(0x20);
+	printk(KERN_DEBUG "... PIC  IRR: %04x\n", v);
+
+	outb(0x0b,0xa0);
+	outb(0x0b,0x20);
+	v = inb(0xa0) << 8 | inb(0x20);
+	outb(0x0a,0xa0);
+	outb(0x0a,0x20);
+
+	spin_unlock_irqrestore(&i8259A_lock, flags);
+
+	printk(KERN_DEBUG "... PIC  ISR: %04x\n", v);
+
+	v = inb(0x4d1) << 8 | inb(0x4d0);
+	printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
+}
+
+__apicdebuginit(int) print_all_ICs(void)
+{
+	print_PIC();
+	print_all_local_APICs();
+	print_IO_APIC();
+
+	return 0;
+}
+
+fs_initcall(print_all_ICs);
+
+
+/* Where if anywhere is the i8259 connect in external int mode */
+static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
+
+void __init enable_IO_APIC(void)
+{
+	union IO_APIC_reg_01 reg_01;
+	int i8259_apic, i8259_pin;
+	int apic;
+	unsigned long flags;
+
+#ifdef CONFIG_X86_32
+	int i;
+	if (!pirqs_enabled)
+		for (i = 0; i < MAX_PIRQS; i++)
+			pirq_entries[i] = -1;
+#endif
+
+	/*
+	 * The number of IO-APIC IRQ registers (== #pins):
+	 */
+	for (apic = 0; apic < nr_ioapics; apic++) {
+		spin_lock_irqsave(&ioapic_lock, flags);
+		reg_01.raw = io_apic_read(apic, 1);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
+	}
+	for(apic = 0; apic < nr_ioapics; apic++) {
+		int pin;
+		/* See if any of the pins is in ExtINT mode */
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+			struct IO_APIC_route_entry entry;
+			entry = ioapic_read_entry(apic, pin);
+
+			/* If the interrupt line is enabled and in ExtInt mode
+			 * I have found the pin where the i8259 is connected.
+			 */
+			if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) {
+				ioapic_i8259.apic = apic;
+				ioapic_i8259.pin  = pin;
+				goto found_i8259;
+			}
+		}
+	}
+ found_i8259:
+	/* Look to see what if the MP table has reported the ExtINT */
+	/* If we could not find the appropriate pin by looking at the ioapic
+	 * the i8259 probably is not connected the ioapic but give the
+	 * mptable a chance anyway.
+	 */
+	i8259_pin  = find_isa_irq_pin(0, mp_ExtINT);
+	i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
+	/* Trust the MP table if nothing is setup in the hardware */
+	if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) {
+		printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n");
+		ioapic_i8259.pin  = i8259_pin;
+		ioapic_i8259.apic = i8259_apic;
+	}
+	/* Complain if the MP table and the hardware disagree */
+	if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) &&
+		(i8259_pin >= 0) && (ioapic_i8259.pin >= 0))
+	{
+		printk(KERN_WARNING "ExtINT in hardware and MP table differ\n");
+	}
+
+	/*
+	 * Do not trust the IO-APIC being empty at bootup
+	 */
+	clear_IO_APIC();
+}
+
+/*
+ * Not an __init, needed by the reboot code
+ */
+void disable_IO_APIC(void)
+{
+	/*
+	 * Clear the IO-APIC before rebooting:
+	 */
+	clear_IO_APIC();
+
+	/*
+	 * 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;
+
+		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            = read_apic_id();
+
+		/*
+		 * Add it to the IO-APIC irq-routing table:
+		 */
+		ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry);
+	}
+
+	disconnect_bsp_APIC(ioapic_i8259.pin != -1);
+}
+
+#ifdef CONFIG_X86_32
+/*
+ * function to set the IO-APIC physical IDs based on the
+ * values stored in the MPC table.
+ *
+ * by Matt Domsch <Matt_Domsch@dell.com>  Tue Dec 21 12:25:05 CST 1999
+ */
+
+static void __init setup_ioapic_ids_from_mpc(void)
+{
+	union IO_APIC_reg_00 reg_00;
+	physid_mask_t phys_id_present_map;
+	int apic;
+	int i;
+	unsigned char old_id;
+	unsigned long flags;
+
+	if (x86_quirks->setup_ioapic_ids && x86_quirks->setup_ioapic_ids())
+		return;
+
+	/*
+	 * Don't check I/O APIC IDs for xAPIC systems.  They have
+	 * no meaning without the serial APIC bus.
+	 */
+	if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
+		|| APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
+		return;
+	/*
+	 * This is broken; anything with a real cpu count has to
+	 * circumvent this idiocy regardless.
+	 */
+	phys_id_present_map = ioapic_phys_id_map(phys_cpu_present_map);
+
+	/*
+	 * 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].mp_apicid;
+
+		if (mp_ioapics[apic].mp_apicid >= get_physical_broadcast()) {
+			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n",
+				apic, mp_ioapics[apic].mp_apicid);
+			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
+				reg_00.bits.ID);
+			mp_ioapics[apic].mp_apicid = reg_00.bits.ID;
+		}
+
+		/*
+		 * Sanity check, is the ID really free? Every APIC in a
+		 * system must have a unique ID or we get lots of nice
+		 * 'stuck on smp_invalidate_needed IPI wait' messages.
+		 */
+		if (check_apicid_used(phys_id_present_map,
+					mp_ioapics[apic].mp_apicid)) {
+			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n",
+				apic, mp_ioapics[apic].mp_apicid);
+			for (i = 0; i < get_physical_broadcast(); i++)
+				if (!physid_isset(i, phys_id_present_map))
+					break;
+			if (i >= get_physical_broadcast())
+				panic("Max APIC ID exceeded!\n");
+			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
+				i);
+			physid_set(i, phys_id_present_map);
+			mp_ioapics[apic].mp_apicid = i;
+		} else {
+			physid_mask_t tmp;
+			tmp = apicid_to_cpu_present(mp_ioapics[apic].mp_apicid);
+			apic_printk(APIC_VERBOSE, "Setting %d in the "
+					"phys_id_present_map\n",
+					mp_ioapics[apic].mp_apicid);
+			physids_or(phys_id_present_map, phys_id_present_map, tmp);
+		}
+
+
+		/*
+		 * We need to adjust the IRQ routing table
+		 * if the ID changed.
+		 */
+		if (old_id != mp_ioapics[apic].mp_apicid)
+			for (i = 0; i < mp_irq_entries; i++)
+				if (mp_irqs[i].mp_dstapic == old_id)
+					mp_irqs[i].mp_dstapic
+						= mp_ioapics[apic].mp_apicid;
+
+		/*
+		 * 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].mp_apicid);
+
+		reg_00.bits.ID = mp_ioapics[apic].mp_apicid;
+		spin_lock_irqsave(&ioapic_lock, flags);
+
+		/*
+		 * Sanity check
+		 */
+		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].mp_apicid)
+			printk("could not set ID!\n");
+		else
+			apic_printk(APIC_VERBOSE, " ok.\n");
+	}
+}
+#endif
+
+int no_timer_check __initdata;
+
+static int __init notimercheck(char *s)
+{
+	no_timer_check = 1;
+	return 1;
+}
+__setup("no_timer_check", notimercheck);
+
+/*
+ * There is a nasty bug in some older SMP boards, their mptable lies
+ * about the timer IRQ. We do the following to work around the situation:
+ *
+ *	- timer IRQ defaults to IO-APIC IRQ
+ *	- if this function detects that timer IRQs are defunct, then we fall
+ *	  back to ISA timer IRQs
+ */
+static int __init timer_irq_works(void)
+{
+	unsigned long t1 = jiffies;
+	unsigned long flags;
+
+	if (no_timer_check)
+		return 1;
+
+	local_save_flags(flags);
+	local_irq_enable();
+	/* Let ten ticks pass... */
+	mdelay((10 * 1000) / HZ);
+	local_irq_restore(flags);
+
+	/*
+	 * Expect a few ticks at least, to be sure some possible
+	 * glue logic does not lock up after one or two first
+	 * ticks in a non-ExtINT mode.  Also the local APIC
+	 * might have cached one ExtINT interrupt.  Finally, at
+	 * least one tick may be lost due to delays.
+	 */
+
+	/* jiffies wrap? */
+	if (time_after(jiffies, t1 + 4))
+		return 1;
+	return 0;
+}
+
+/*
+ * In the SMP+IOAPIC case it might happen that there are an unspecified
+ * number of pending IRQ events unhandled. These cases are very rare,
+ * so we 'resend' these IRQs via IPIs, to the same CPU. It's much
+ * better to do it this way as thus we do not have to be aware of
+ * 'pending' interrupts in the IRQ path, except at this point.
+ */
+/*
+ * Edge triggered needs to resend any interrupt
+ * that was delayed but this is now handled in the device
+ * independent code.
+ */
+
+/*
+ * Starting up a edge-triggered IO-APIC interrupt is
+ * nasty - we need to make sure that we get the edge.
+ * If it is already asserted for some reason, we need
+ * return 1 to indicate that is was pending.
+ *
+ * This is not complete - we should be able to fake
+ * an edge even if it isn't on the 8259A...
+ */
+
+static unsigned int startup_ioapic_irq(unsigned int irq)
+{
+	int was_pending = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	if (irq < 16) {
+		disable_8259A_irq(irq);
+		if (i8259A_irq_pending(irq))
+			was_pending = 1;
+	}
+	__unmask_IO_APIC_irq(irq);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	return was_pending;
+}
+
+#ifdef CONFIG_X86_64
+static int ioapic_retrigger_irq(unsigned int irq)
+{
+
+	struct irq_cfg *cfg = irq_cfg(irq);
+	unsigned long flags;
+
+	spin_lock_irqsave(&vector_lock, flags);
+	send_IPI_mask(cpumask_of_cpu(first_cpu(cfg->domain)), cfg->vector);
+	spin_unlock_irqrestore(&vector_lock, flags);
+
+	return 1;
+}
+#else
+static int ioapic_retrigger_irq(unsigned int irq)
+{
+        send_IPI_self(irq_cfg(irq)->vector);
+
+        return 1;
+}
+#endif
+
+/*
+ * Level and edge triggered IO-APIC interrupts need different handling,
+ * so we use two separate IRQ descriptors. Edge triggered IRQs can be
+ * handled with the level-triggered descriptor, but that one has slightly
+ * more overhead. Level-triggered interrupts cannot be handled with the
+ * edge-triggered handler, without risking IRQ storms and other ugly
+ * races.
+ */
+
+#ifdef CONFIG_SMP
+
+#ifdef CONFIG_INTR_REMAP
+static void ir_irq_migration(struct work_struct *work);
+
+static DECLARE_DELAYED_WORK(ir_migration_work, ir_irq_migration);
+
+/*
+ * Migrate the IO-APIC irq in the presence of intr-remapping.
+ *
+ * For edge triggered, irq migration is a simple atomic update(of vector
+ * and cpu destination) of IRTE and flush the hardware cache.
+ *
+ * For level triggered, we need to modify the io-apic RTE aswell with the update
+ * vector information, along with modifying IRTE with vector and destination.
+ * So irq migration for level triggered is little  bit more complex compared to
+ * edge triggered migration. But the good news is, we use the same algorithm
+ * for level triggered migration as we have today, only difference being,
+ * we now initiate the irq migration from process context instead of the
+ * interrupt context.
+ *
+ * In future, when we do a directed EOI (combined with cpu EOI broadcast
+ * suppression) to the IO-APIC, level triggered irq migration will also be
+ * as simple as edge triggered migration and we can do the irq migration
+ * with a simple atomic update to IO-APIC RTE.
+ */
+static void migrate_ioapic_irq(int irq, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	struct irq_desc *desc;
+	cpumask_t tmp, cleanup_mask;
+	struct irte irte;
+	int modify_ioapic_rte;
+	unsigned int dest;
+	unsigned long flags;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (get_irte(irq, &irte))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	desc = irq_desc(irq);
+	modify_ioapic_rte = desc->status & IRQ_LEVEL;
+	if (modify_ioapic_rte) {
+		spin_lock_irqsave(&ioapic_lock, flags);
+		__target_IO_APIC_irq(irq, dest, cfg->vector);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+	}
+
+	irte.vector = cfg->vector;
+	irte.dest_id = IRTE_DEST(dest);
+
+	/*
+	 * Modified the IRTE and flushes the Interrupt entry cache.
+	 */
+	modify_irte(irq, &irte);
+
+	if (cfg->move_in_progress) {
+		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
+		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
+		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+		cfg->move_in_progress = 0;
+	}
+
+	desc->affinity = mask;
+}
+
+static int migrate_irq_remapped_level(int irq)
+{
+	int ret = -1;
+	struct irq_desc *desc = irq_desc(irq);
+
+	mask_IO_APIC_irq(irq);
+
+	if (io_apic_level_ack_pending(irq)) {
+		/*
+	 	 * Interrupt in progress. Migrating irq now will change the
+		 * vector information in the IO-APIC RTE and that will confuse
+		 * the EOI broadcast performed by cpu.
+		 * So, delay the irq migration to the next instance.
+		 */
+		schedule_delayed_work(&ir_migration_work, 1);
+		goto unmask;
+	}
+
+	/* everthing is clear. we have right of way */
+	migrate_ioapic_irq(irq, desc->pending_mask);
+
+	ret = 0;
+	desc->status &= ~IRQ_MOVE_PENDING;
+	cpus_clear(desc->pending_mask);
+
+unmask:
+	unmask_IO_APIC_irq(irq);
+	return ret;
+}
+
+static void ir_irq_migration(struct work_struct *work)
+{
+	unsigned int irq;
+	struct irq_desc *desc;
+
+	for_each_irq_desc(irq, desc) {
+		if (desc->status & IRQ_MOVE_PENDING) {
+			unsigned long flags;
+
+			spin_lock_irqsave(&desc->lock, flags);
+			if (!desc->chip->set_affinity ||
+			    !(desc->status & IRQ_MOVE_PENDING)) {
+				desc->status &= ~IRQ_MOVE_PENDING;
+				spin_unlock_irqrestore(&desc->lock, flags);
+				continue;
+			}
+
+			desc->chip->set_affinity(irq, desc, desc->pending_mask);
+			spin_unlock_irqrestore(&desc->lock, flags);
+		}
+	}
+}
+
+/*
+ * Migrates the IRQ destination in the process context.
+ */
+static void set_ir_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	if (desc->status & IRQ_LEVEL) {
+		desc->status |= IRQ_MOVE_PENDING;
+		desc->pending_mask = mask;
+		migrate_irq_remapped_level(irq);
+		return;
+	}
+
+	migrate_ioapic_irq(irq, mask);
+}
+#endif
+
+asmlinkage void smp_irq_move_cleanup_interrupt(void)
+{
+	unsigned vector, me;
+	ack_APIC_irq();
+#ifdef CONFIG_X86_64
+	exit_idle();
+#endif
+	irq_enter();
+
+	me = smp_processor_id();
+	for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
+		unsigned int irq;
+		struct irq_desc *desc;
+		struct irq_cfg *cfg;
+		irq = __get_cpu_var(vector_irq)[vector];
+
+		desc = irq_desc(irq);
+		if (!desc)
+			continue;
+
+		cfg = irq_cfg(irq);
+		spin_lock(&desc->lock);
+		if (!cfg->move_cleanup_count)
+			goto unlock;
+
+		if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
+			goto unlock;
+
+		__get_cpu_var(vector_irq)[vector] = -1;
+		cfg->move_cleanup_count--;
+unlock:
+		spin_unlock(&desc->lock);
+	}
+
+	irq_exit();
+}
+
+static void irq_complete_move(unsigned int irq)
+{
+	struct irq_cfg *cfg = irq_cfg(irq);
+	unsigned vector, me;
+
+	if (likely(!cfg->move_in_progress))
+		return;
+
+	vector = ~get_irq_regs()->orig_ax;
+	me = smp_processor_id();
+	if ((vector == cfg->vector) && cpu_isset(me, cfg->domain)) {
+		cpumask_t cleanup_mask;
+
+		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
+		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
+		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+		cfg->move_in_progress = 0;
+	}
+}
+#else
+static inline void irq_complete_move(unsigned int irq) {}
+#endif
+#ifdef CONFIG_INTR_REMAP
+static void ack_x2apic_level(unsigned int irq, struct irq_desc *desc)
+{
+	ack_x2APIC_irq();
+}
+
+static void ack_x2apic_edge(unsigned int irq, struct irq_desc *desc)
+{
+	ack_x2APIC_irq();
+}
+#endif
+
+static void ack_apic_edge(unsigned int irq, struct irq_desc *desc)
+{
+	irq_complete_move(irq);
+	move_native_irq(irq, desc);
+	ack_APIC_irq();
+}
+
+#ifdef CONFIG_X86_64
+static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
+{
+	int do_unmask_irq = 0;
+
+	irq_complete_move(irq);
+#ifdef CONFIG_GENERIC_PENDING_IRQ
+	/* If we are moving the irq we need to mask it */
+	if (unlikely(desc->status & IRQ_MOVE_PENDING)) {
+		do_unmask_irq = 1;
+		mask_IO_APIC_irq(irq);
+	}
+#endif
+
+	/*
+	 * We must acknowledge the irq before we move it or the acknowledge will
+	 * not propagate properly.
+	 */
+	ack_APIC_irq();
+
+	/* Now we can move and renable the irq */
+	if (unlikely(do_unmask_irq)) {
+		/* Only migrate the irq if the ack has been received.
+		 *
+		 * On rare occasions the broadcast level triggered ack gets
+		 * delayed going to ioapics, and if we reprogram the
+		 * vector while Remote IRR is still set the irq will never
+		 * fire again.
+		 *
+		 * To prevent this scenario we read the Remote IRR bit
+		 * of the ioapic.  This has two effects.
+		 * - On any sane system the read of the ioapic will
+		 *   flush writes (and acks) going to the ioapic from
+		 *   this cpu.
+		 * - We get to see if the ACK has actually been delivered.
+		 *
+		 * Based on failed experiments of reprogramming the
+		 * ioapic entry from outside of irq context starting
+		 * with masking the ioapic entry and then polling until
+		 * Remote IRR was clear before reprogramming the
+		 * ioapic I don't trust the Remote IRR bit to be
+		 * completey accurate.
+		 *
+		 * However there appears to be no other way to plug
+		 * this race, so if the Remote IRR bit is not
+		 * accurate and is causing problems then it is a hardware bug
+		 * and you can go talk to the chipset vendor about it.
+		 */
+		if (!io_apic_level_ack_pending(irq))
+			move_masked_irq(irq, desc);
+		unmask_IO_APIC_irq(irq);
+	}
+}
+#else
+atomic_t irq_mis_count;
+static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
+{
+	unsigned long v;
+	int i;
+
+	irq_complete_move(irq);
+	move_native_irq(irq, desc);
+	/*
+	* 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
+	*/
+	i = irq_cfg(irq)->vector;
+
+	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
+
+	ack_APIC_irq();
+
+	if (!(v & (1 << (i & 0x1f)))) {
+		atomic_inc(&irq_mis_count);
+		spin_lock(&ioapic_lock);
+		__mask_and_edge_IO_APIC_irq(irq);
+		__unmask_and_level_IO_APIC_irq(irq);
+		spin_unlock(&ioapic_lock);
+	}
+}
+#endif
+
+static struct irq_chip ioapic_chip __read_mostly = {
+	.name 		= "IO-APIC",
+	.startup 	= startup_ioapic_irq,
+	.mask	 	= mask_IO_APIC_irq,
+	.unmask	 	= unmask_IO_APIC_irq,
+	.ack 		= ack_apic_edge,
+	.eoi 		= ack_apic_level,
+#ifdef CONFIG_SMP
+	.set_affinity 	= set_ioapic_affinity_irq,
+#endif
+	.retrigger	= ioapic_retrigger_irq,
+};
+
+#ifdef CONFIG_INTR_REMAP
+static struct irq_chip ir_ioapic_chip __read_mostly = {
+	.name 		= "IR-IO-APIC",
+	.startup 	= startup_ioapic_irq,
+	.mask	 	= mask_IO_APIC_irq,
+	.unmask	 	= unmask_IO_APIC_irq,
+	.ack 		= ack_x2apic_edge,
+	.eoi 		= ack_x2apic_level,
+#ifdef CONFIG_SMP
+	.set_affinity 	= set_ir_ioapic_affinity_irq,
+#endif
+	.retrigger	= ioapic_retrigger_irq,
+};
+#endif
+
+static inline void init_IO_APIC_traps(void)
+{
+	int irq;
+	struct irq_desc *desc;
+	struct irq_cfg *cfg;
+
+	/*
+	 * NOTE! The local APIC isn't very good at handling
+	 * multiple interrupts at the same interrupt level.
+	 * As the interrupt level is determined by taking the
+	 * vector number and shifting that right by 4, we
+	 * want to spread these out a bit so that they don't
+	 * all fall in the same interrupt level.
+	 *
+	 * Also, we've got to be careful not to trash gate
+	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
+	 */
+	for_each_irq_cfg(cfg) {
+		irq = cfg->irq;
+		if (IO_APIC_IRQ(irq) && !cfg->vector) {
+			/*
+			 * Hmm.. We don't have an entry for this,
+			 * so default to an old-fashioned 8259
+			 * interrupt if we can..
+			 */
+			if (irq < 16)
+				make_8259A_irq(irq);
+			else {
+				desc = irq_desc(irq);
+				/* Strange. Oh, well.. */
+				desc->chip = &no_irq_chip;
+			}
+		}
+	}
+}
+
+/*
+ * The local APIC irq-chip implementation:
+ */
+
+static void mask_lapic_irq(unsigned int irq)
+{
+	unsigned long v;
+
+	v = apic_read(APIC_LVT0);
+	apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
+}
+
+static void unmask_lapic_irq(unsigned int irq)
+{
+	unsigned long v;
+
+	v = apic_read(APIC_LVT0);
+	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
+}
+
+static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
+{
+	ack_APIC_irq();
+}
+
+static struct irq_chip lapic_chip __read_mostly = {
+	.name		= "local-APIC",
+	.mask		= mask_lapic_irq,
+	.unmask		= unmask_lapic_irq,
+	.ack		= ack_lapic_irq,
+};
+
+static void lapic_register_intr(int irq)
+{
+	struct irq_desc *desc;
+
+	desc = irq_desc(irq);
+	desc->status &= ~IRQ_LEVEL;
+	set_irq_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq,
+				      "edge");
+}
+
+static void __init setup_nmi(void)
+{
+	/*
+	 * Dirty trick to enable the NMI watchdog ...
+	 * We put the 8259A master into AEOI mode and
+	 * unmask on all local APICs LVT0 as NMI.
+	 *
+	 * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
+	 * is from Maciej W. Rozycki - so we do not have to EOI from
+	 * the NMI handler or the timer interrupt.
+	 */
+	apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ...");
+
+	enable_NMI_through_LVT0();
+
+	apic_printk(APIC_VERBOSE, " done.\n");
+}
+
+/*
+ * This looks a bit hackish but it's about the only one way of sending
+ * a few INTA cycles to 8259As and any associated glue logic.  ICR does
+ * not support the ExtINT mode, unfortunately.  We need to send these
+ * cycles as some i82489DX-based boards have glue logic that keeps the
+ * 8259A interrupt line asserted until INTA.  --macro
+ */
+static inline void __init unlock_ExtINT_logic(void)
+{
+	int apic, pin, i;
+	struct IO_APIC_route_entry entry0, entry1;
+	unsigned char save_control, save_freq_select;
+
+	pin  = find_isa_irq_pin(8, mp_INT);
+	if (pin == -1) {
+		WARN_ON_ONCE(1);
+		return;
+	}
+	apic = find_isa_irq_apic(8, mp_INT);
+	if (apic == -1) {
+		WARN_ON_ONCE(1);
+		return;
+	}
+
+	entry0 = ioapic_read_entry(apic, pin);
+	clear_IO_APIC_pin(apic, pin);
+
+	memset(&entry1, 0, sizeof(entry1));
+
+	entry1.dest_mode = 0;			/* physical delivery */
+	entry1.mask = 0;			/* unmask IRQ now */
+	entry1.dest = hard_smp_processor_id();
+	entry1.delivery_mode = dest_ExtINT;
+	entry1.polarity = entry0.polarity;
+	entry1.trigger = 0;
+	entry1.vector = 0;
+
+	ioapic_write_entry(apic, pin, entry1);
+
+	save_control = CMOS_READ(RTC_CONTROL);
+	save_freq_select = CMOS_READ(RTC_FREQ_SELECT);
+	CMOS_WRITE((save_freq_select & ~RTC_RATE_SELECT) | 0x6,
+		   RTC_FREQ_SELECT);
+	CMOS_WRITE(save_control | RTC_PIE, RTC_CONTROL);
+
+	i = 100;
+	while (i-- > 0) {
+		mdelay(10);
+		if ((CMOS_READ(RTC_INTR_FLAGS) & RTC_PF) == RTC_PF)
+			i -= 10;
+	}
+
+	CMOS_WRITE(save_control, RTC_CONTROL);
+	CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
+	clear_IO_APIC_pin(apic, pin);
+
+	ioapic_write_entry(apic, pin, entry0);
+}
+
+static int disable_timer_pin_1 __initdata;
+/* Actually the next is obsolete, but keep it for paranoid reasons -AK */
+static int __init disable_timer_pin_setup(char *arg)
+{
+	disable_timer_pin_1 = 1;
+	return 0;
+}
+early_param("disable_timer_pin_1", disable_timer_pin_setup);
+
+int timer_through_8259 __initdata;
+
+/*
+ * This code may look a bit paranoid, but it's supposed to cooperate with
+ * a wide range of boards and BIOS bugs.  Fortunately only the timer IRQ
+ * is so screwy.  Thanks to Brian Perkins for testing/hacking this beast
+ * fanatically on his truly buggy board.
+ *
+ * FIXME: really need to revamp this for all platforms.
+ */
+static inline void __init check_timer(void)
+{
+	struct irq_cfg *cfg = irq_cfg(0);
+	int apic1, pin1, apic2, pin2;
+	unsigned long flags;
+	unsigned int ver;
+	int no_pin1 = 0;
+
+	local_irq_save(flags);
+
+        ver = apic_read(APIC_LVR);
+        ver = GET_APIC_VERSION(ver);
+
+	/*
+	 * get/set the timer IRQ vector:
+	 */
+	disable_8259A_irq(0);
+	assign_irq_vector(0, TARGET_CPUS);
+
+	/*
+	 * As IRQ0 is to be enabled in the 8259A, the virtual
+	 * wire has to be disabled in the local APIC.  Also
+	 * timer interrupts need to be acknowledged manually in
+	 * the 8259A for the i82489DX when using the NMI
+	 * watchdog as that APIC treats NMIs as level-triggered.
+	 * The AEOI mode will finish them in the 8259A
+	 * automatically.
+	 */
+	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
+	init_8259A(1);
+#ifdef CONFIG_X86_32
+	timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver));
+#endif
+
+	pin1  = find_isa_irq_pin(0, mp_INT);
+	apic1 = find_isa_irq_apic(0, mp_INT);
+	pin2  = ioapic_i8259.pin;
+	apic2 = ioapic_i8259.apic;
+
+	apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X "
+		    "apic1=%d pin1=%d apic2=%d pin2=%d\n",
+		    cfg->vector, apic1, pin1, apic2, pin2);
+
+	/*
+	 * Some BIOS writers are clueless and report the ExtINTA
+	 * I/O APIC input from the cascaded 8259A as the timer
+	 * interrupt input.  So just in case, if only one pin
+	 * was found above, try it both directly and through the
+	 * 8259A.
+	 */
+	if (pin1 == -1) {
+#ifdef CONFIG_INTR_REMAP
+		if (intr_remapping_enabled)
+			panic("BIOS bug: timer not connected to IO-APIC");
+#endif
+		pin1 = pin2;
+		apic1 = apic2;
+		no_pin1 = 1;
+	} else if (pin2 == -1) {
+		pin2 = pin1;
+		apic2 = apic1;
+	}
+
+	if (pin1 != -1) {
+		/*
+		 * Ok, does IRQ0 through the IOAPIC work?
+		 */
+		if (no_pin1) {
+			add_pin_to_irq(0, apic1, pin1);
+			setup_timer_IRQ0_pin(apic1, pin1, cfg->vector);
+		}
+		unmask_IO_APIC_irq(0);
+		if (timer_irq_works()) {
+			if (nmi_watchdog == NMI_IO_APIC) {
+				setup_nmi();
+				enable_8259A_irq(0);
+			}
+			if (disable_timer_pin_1 > 0)
+				clear_IO_APIC_pin(0, pin1);
+			goto out;
+		}
+#ifdef CONFIG_INTR_REMAP
+		if (intr_remapping_enabled)
+			panic("timer doesn't work through Interrupt-remapped IO-APIC");
+#endif
+		clear_IO_APIC_pin(apic1, pin1);
+		if (!no_pin1)
+			apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
+				    "8254 timer not connected to IO-APIC\n");
+
+		apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer "
+			    "(IRQ0) through the 8259A ...\n");
+		apic_printk(APIC_QUIET, KERN_INFO
+			    "..... (found apic %d pin %d) ...\n", apic2, pin2);
+		/*
+		 * legacy devices should be connected to IO APIC #0
+		 */
+		replace_pin_at_irq(0, apic1, pin1, apic2, pin2);
+		setup_timer_IRQ0_pin(apic2, pin2, cfg->vector);
+		unmask_IO_APIC_irq(0);
+		enable_8259A_irq(0);
+		if (timer_irq_works()) {
+			apic_printk(APIC_QUIET, KERN_INFO "....... works.\n");
+			timer_through_8259 = 1;
+			if (nmi_watchdog == NMI_IO_APIC) {
+				disable_8259A_irq(0);
+				setup_nmi();
+				enable_8259A_irq(0);
+			}
+			goto out;
+		}
+		/*
+		 * Cleanup, just in case ...
+		 */
+		disable_8259A_irq(0);
+		clear_IO_APIC_pin(apic2, pin2);
+		apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n");
+	}
+
+	if (nmi_watchdog == NMI_IO_APIC) {
+		apic_printk(APIC_QUIET, KERN_WARNING "timer doesn't work "
+			    "through the IO-APIC - disabling NMI Watchdog!\n");
+		nmi_watchdog = NMI_NONE;
+	}
+#ifdef CONFIG_X86_32
+	timer_ack = 0;
+#endif
+
+	apic_printk(APIC_QUIET, KERN_INFO
+		    "...trying to set up timer as Virtual Wire IRQ...\n");
+
+	lapic_register_intr(0);
+	apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector);	/* Fixed mode */
+	enable_8259A_irq(0);
+
+	if (timer_irq_works()) {
+		apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
+		goto out;
+	}
+	disable_8259A_irq(0);
+	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector);
+	apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n");
+
+	apic_printk(APIC_QUIET, KERN_INFO
+		    "...trying to set up timer as ExtINT IRQ...\n");
+
+	init_8259A(0);
+	make_8259A_irq(0);
+	apic_write(APIC_LVT0, APIC_DM_EXTINT);
+
+	unlock_ExtINT_logic();
+
+	if (timer_irq_works()) {
+		apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
+		goto out;
+	}
+	apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n");
+	panic("IO-APIC + timer doesn't work!  Boot with apic=debug and send a "
+		"report.  Then try booting with the 'noapic' option.\n");
+out:
+	local_irq_restore(flags);
+}
+
+/*
+ * Traditionally ISA IRQ2 is the cascade IRQ, and is not available
+ * to devices.  However there may be an I/O APIC pin available for
+ * this interrupt regardless.  The pin may be left unconnected, but
+ * typically it will be reused as an ExtINT cascade interrupt for
+ * the master 8259A.  In the MPS case such a pin will normally be
+ * reported as an ExtINT interrupt in the MP table.  With ACPI
+ * there is no provision for ExtINT interrupts, and in the absence
+ * of an override it would be treated as an ordinary ISA I/O APIC
+ * interrupt, that is edge-triggered and unmasked by default.  We
+ * used to do this, but it caused problems on some systems because
+ * of the NMI watchdog and sometimes IRQ0 of the 8254 timer using
+ * the same ExtINT cascade interrupt to drive the local APIC of the
+ * bootstrap processor.  Therefore we refrain from routing IRQ2 to
+ * the I/O APIC in all cases now.  No actual device should request
+ * it anyway.  --macro
+ */
+#define PIC_IRQS	(1 << PIC_CASCADE_IR)
+
+void __init setup_IO_APIC(void)
+{
+
+#ifdef CONFIG_X86_32
+	enable_IO_APIC();
+#else
+	/*
+	 * calling enable_IO_APIC() is moved to setup_local_APIC for BP
+	 */
+#endif
+
+	io_apic_irqs = ~PIC_IRQS;
+
+	apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
+        /*
+         * Set up IO-APIC IRQ routing.
+         */
+#ifdef CONFIG_X86_32
+        if (!acpi_ioapic)
+                setup_ioapic_ids_from_mpc();
+#endif
+	sync_Arb_IDs();
+	setup_IO_APIC_irqs();
+	init_IO_APIC_traps();
+	check_timer();
+}
+
+/*
+ *      Called after all the initialization is done. If we didnt find any
+ *      APIC bugs then we can allow the modify fast path
+ */
+
+static int __init io_apic_bug_finalize(void)
+{
+        if (sis_apic_bug == -1)
+                sis_apic_bug = 0;
+        return 0;
+}
+
+late_initcall(io_apic_bug_finalize);
+
+struct sysfs_ioapic_data {
+	struct sys_device dev;
+	struct IO_APIC_route_entry entry[0];
+};
+static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
+
+static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
+{
+	struct IO_APIC_route_entry *entry;
+	struct sysfs_ioapic_data *data;
+	int i;
+
+	data = container_of(dev, struct sysfs_ioapic_data, dev);
+	entry = data->entry;
+	for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ )
+		*entry = ioapic_read_entry(dev->id, i);
+
+	return 0;
+}
+
+static int ioapic_resume(struct sys_device *dev)
+{
+	struct IO_APIC_route_entry *entry;
+	struct sysfs_ioapic_data *data;
+	unsigned long flags;
+	union IO_APIC_reg_00 reg_00;
+	int i;
+
+	data = container_of(dev, struct sysfs_ioapic_data, dev);
+	entry = data->entry;
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	reg_00.raw = io_apic_read(dev->id, 0);
+	if (reg_00.bits.ID != mp_ioapics[dev->id].mp_apicid) {
+		reg_00.bits.ID = mp_ioapics[dev->id].mp_apicid;
+		io_apic_write(dev->id, 0, reg_00.raw);
+	}
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+	for (i = 0; i < nr_ioapic_registers[dev->id]; i++)
+		ioapic_write_entry(dev->id, i, entry[i]);
+
+	return 0;
+}
+
+static struct sysdev_class ioapic_sysdev_class = {
+	.name = "ioapic",
+	.suspend = ioapic_suspend,
+	.resume = ioapic_resume,
+};
+
+static int __init ioapic_init_sysfs(void)
+{
+	struct sys_device * dev;
+	int i, size, error;
+
+	error = sysdev_class_register(&ioapic_sysdev_class);
+	if (error)
+		return error;
+
+	for (i = 0; i < nr_ioapics; i++ ) {
+		size = sizeof(struct sys_device) + nr_ioapic_registers[i]
+			* sizeof(struct IO_APIC_route_entry);
+		mp_ioapic_data[i] = kzalloc(size, GFP_KERNEL);
+		if (!mp_ioapic_data[i]) {
+			printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
+			continue;
+		}
+		dev = &mp_ioapic_data[i]->dev;
+		dev->id = i;
+		dev->cls = &ioapic_sysdev_class;
+		error = sysdev_register(dev);
+		if (error) {
+			kfree(mp_ioapic_data[i]);
+			mp_ioapic_data[i] = NULL;
+			printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
+			continue;
+		}
+	}
+
+	return 0;
+}
+
+device_initcall(ioapic_init_sysfs);
+
+/*
+ * Dynamic irq allocate and deallocation
+ */
+unsigned int create_irq(unsigned int irq_want)
+{
+	/* Allocate an unused irq */
+	unsigned int irq;
+	unsigned int new;
+	unsigned long flags;
+	struct irq_cfg *cfg_new;
+
+#ifndef CONFIG_HAVE_SPARSE_IRQ
+	irq_want = nr_irqs - 1;
+#endif
+
+	irq = 0;
+	spin_lock_irqsave(&vector_lock, flags);
+	for (new = irq_want; new > 0; new--) {
+		if (platform_legacy_irq(new))
+			continue;
+		cfg_new = irq_cfg(new);
+		if (cfg_new && cfg_new->vector != 0)
+			continue;
+		/* check if need to create one */
+		if (!cfg_new)
+			cfg_new = irq_cfg_with_new(new);
+		if (__assign_irq_vector(new, TARGET_CPUS) == 0)
+			irq = new;
+		break;
+	}
+	spin_unlock_irqrestore(&vector_lock, flags);
+
+	if (irq > 0) {
+		dynamic_irq_init(irq);
+	}
+	return irq;
+}
+
+void destroy_irq(unsigned int irq)
+{
+	unsigned long flags;
+
+	dynamic_irq_cleanup(irq);
+
+#ifdef CONFIG_INTR_REMAP
+	free_irte(irq);
+#endif
+	spin_lock_irqsave(&vector_lock, flags);
+	__clear_irq_vector(irq);
+	spin_unlock_irqrestore(&vector_lock, flags);
+}
+
+/*
+ * MSI message composition
+ */
+#ifdef CONFIG_PCI_MSI
+static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
+{
+	struct irq_cfg *cfg;
+	int err;
+	unsigned dest;
+	cpumask_t tmp;
+
+	tmp = TARGET_CPUS;
+	err = assign_irq_vector(irq, tmp);
+	if (err)
+		return err;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, tmp);
+	dest = cpu_mask_to_apicid(tmp);
+
+#ifdef CONFIG_INTR_REMAP
+	if (irq_remapped(irq)) {
+		struct irte irte;
+		int ir_index;
+		u16 sub_handle;
+
+		ir_index = map_irq_to_irte_handle(irq, &sub_handle);
+		BUG_ON(ir_index == -1);
+
+		memset (&irte, 0, sizeof(irte));
+
+		irte.present = 1;
+		irte.dst_mode = INT_DEST_MODE;
+		irte.trigger_mode = 0; /* edge */
+		irte.dlvry_mode = INT_DELIVERY_MODE;
+		irte.vector = cfg->vector;
+		irte.dest_id = IRTE_DEST(dest);
+
+		modify_irte(irq, &irte);
+
+		msg->address_hi = MSI_ADDR_BASE_HI;
+		msg->data = sub_handle;
+		msg->address_lo = MSI_ADDR_BASE_LO | MSI_ADDR_IR_EXT_INT |
+				  MSI_ADDR_IR_SHV |
+				  MSI_ADDR_IR_INDEX1(ir_index) |
+				  MSI_ADDR_IR_INDEX2(ir_index);
+	} else
+#endif
+	{
+		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(cfg->vector);
+	}
+	return err;
+}
+
+#ifdef CONFIG_SMP
+static void set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	struct msi_msg msg;
+	unsigned int dest;
+	cpumask_t tmp;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	read_msi_msg(irq, &msg);
+
+	msg.data &= ~MSI_DATA_VECTOR_MASK;
+	msg.data |= MSI_DATA_VECTOR(cfg->vector);
+	msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
+	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
+
+	write_msi_msg(irq, &msg);
+	desc->affinity = mask;
+}
+
+#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 ir_set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	unsigned int dest;
+	cpumask_t tmp, cleanup_mask;
+	struct irte irte;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (get_irte(irq, &irte))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	irte.vector = cfg->vector;
+	irte.dest_id = IRTE_DEST(dest);
+
+	/*
+	 * atomically update the IRTE with the new destination and vector.
+	 */
+	modify_irte(irq, &irte);
+
+	/*
+	 * After this point, all the interrupts will start arriving
+	 * at the new destination. So, time to cleanup the previous
+	 * vector allocation.
+	 */
+	if (cfg->move_in_progress) {
+		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
+		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
+		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+		cfg->move_in_progress = 0;
+	}
+
+	desc->affinity = mask;
+}
+#endif
+#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,
+};
+
+#ifdef CONFIG_INTR_REMAP
+static struct irq_chip msi_ir_chip = {
+	.name		= "IR-PCI-MSI",
+	.unmask		= unmask_msi_irq,
+	.mask		= mask_msi_irq,
+	.ack		= ack_x2apic_edge,
+#ifdef CONFIG_SMP
+	.set_affinity	= ir_set_msi_irq_affinity,
+#endif
+	.retrigger	= ioapic_retrigger_irq,
+};
+
+/*
+ * Map the PCI dev to the corresponding remapping hardware unit
+ * and allocate 'nvec' consecutive interrupt-remapping table entries
+ * in it.
+ */
+static int msi_alloc_irte(struct pci_dev *dev, int irq, int nvec)
+{
+	struct intel_iommu *iommu;
+	int index;
+
+	iommu = map_dev_to_ir(dev);
+	if (!iommu) {
+		printk(KERN_ERR
+		       "Unable to map PCI %s to iommu\n", pci_name(dev));
+		return -ENOENT;
+	}
+
+	index = alloc_irte(iommu, irq, nvec);
+	if (index < 0) {
+		printk(KERN_ERR
+		       "Unable to allocate %d IRTE for PCI %s\n", nvec,
+		        pci_name(dev));
+		return -ENOSPC;
+	}
+	return index;
+}
+#endif
+
+static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc, int irq)
+{
+	int ret;
+	struct msi_msg msg;
+
+	ret = msi_compose_msg(dev, irq, &msg);
+	if (ret < 0)
+		return ret;
+
+	set_irq_msi(irq, desc);
+	write_msi_msg(irq, &msg);
+
+#ifdef CONFIG_INTR_REMAP
+	if (irq_remapped(irq)) {
+		struct irq_desc *desc = irq_desc(irq);
+		/*
+		 * irq migration in process context
+		 */
+		desc->status |= IRQ_MOVE_PCNTXT;
+		set_irq_chip_and_handler_name(irq, &msi_ir_chip, handle_edge_irq, "edge");
+	} else
+#endif
+		set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
+
+	return 0;
+}
+
+static unsigned int build_irq_for_pci_dev(struct pci_dev *dev)
+{
+	unsigned int irq;
+
+	irq = dev->bus->number;
+	irq <<= 8;
+	irq |= dev->devfn;
+	irq <<= 12;
+
+	return irq;
+}
+
+int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
+{
+	unsigned int irq;
+	int ret;
+	unsigned int irq_want;
+
+	irq_want = build_irq_for_pci_dev(dev) + 0x100;
+
+	irq = create_irq(irq_want);
+	if (irq == 0)
+		return -1;
+
+#ifdef CONFIG_INTR_REMAP
+	if (!intr_remapping_enabled)
+		goto no_ir;
+
+	ret = msi_alloc_irte(dev, irq, 1);
+	if (ret < 0)
+		goto error;
+no_ir:
+#endif
+	ret = setup_msi_irq(dev, desc, irq);
+	if (ret < 0) {
+		destroy_irq(irq);
+		return ret;
+	}
+	return 0;
+
+#ifdef CONFIG_INTR_REMAP
+error:
+	destroy_irq(irq);
+	return ret;
+#endif
+}
+
+int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
+{
+	unsigned int irq;
+	int ret, sub_handle;
+	struct msi_desc *desc;
+	unsigned int irq_want;
+
+#ifdef CONFIG_INTR_REMAP
+	struct intel_iommu *iommu = 0;
+	int index = 0;
+#endif
+
+	irq_want = build_irq_for_pci_dev(dev) + 0x100;
+	sub_handle = 0;
+	list_for_each_entry(desc, &dev->msi_list, list) {
+		irq = create_irq(irq_want--);
+		if (irq == 0)
+			return -1;
+#ifdef CONFIG_INTR_REMAP
+		if (!intr_remapping_enabled)
+			goto no_ir;
+
+		if (!sub_handle) {
+			/*
+			 * allocate the consecutive block of IRTE's
+			 * for 'nvec'
+			 */
+			index = msi_alloc_irte(dev, irq, nvec);
+			if (index < 0) {
+				ret = index;
+				goto error;
+			}
+		} else {
+			iommu = map_dev_to_ir(dev);
+			if (!iommu) {
+				ret = -ENOENT;
+				goto error;
+			}
+			/*
+			 * setup the mapping between the irq and the IRTE
+			 * base index, the sub_handle pointing to the
+			 * appropriate interrupt remap table entry.
+			 */
+			set_irte_irq(irq, iommu, index, sub_handle);
+		}
+no_ir:
+#endif
+		ret = setup_msi_irq(dev, desc, irq);
+		if (ret < 0)
+			goto error;
+		sub_handle++;
+	}
+	return 0;
+
+error:
+	destroy_irq(irq);
+	return ret;
+}
+
+void arch_teardown_msi_irq(unsigned int irq)
+{
+	destroy_irq(irq);
+}
+
+#ifdef CONFIG_DMAR
+#ifdef CONFIG_SMP
+static void dmar_msi_set_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	struct msi_msg msg;
+	unsigned int dest;
+	cpumask_t tmp;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	dmar_msi_read(irq, &msg);
+
+	msg.data &= ~MSI_DATA_VECTOR_MASK;
+	msg.data |= MSI_DATA_VECTOR(cfg->vector);
+	msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
+	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
+
+	dmar_msi_write(irq, &msg);
+	desc->affinity = mask;
+}
+#endif /* CONFIG_SMP */
+
+struct irq_chip dmar_msi_type = {
+	.name = "DMAR_MSI",
+	.unmask = dmar_msi_unmask,
+	.mask = dmar_msi_mask,
+	.ack = ack_apic_edge,
+#ifdef CONFIG_SMP
+	.set_affinity = dmar_msi_set_affinity,
+#endif
+	.retrigger = ioapic_retrigger_irq,
+};
+
+int arch_setup_dmar_msi(unsigned int irq)
+{
+	int ret;
+	struct msi_msg msg;
+
+	ret = msi_compose_msg(NULL, irq, &msg);
+	if (ret < 0)
+		return ret;
+	dmar_msi_write(irq, &msg);
+	set_irq_chip_and_handler_name(irq, &dmar_msi_type, handle_edge_irq,
+		"edge");
+	return 0;
+}
+#endif
+
+#endif /* CONFIG_PCI_MSI */
+/*
+ * Hypertransport interrupt support
+ */
+#ifdef CONFIG_HT_IRQ
+
+#ifdef CONFIG_SMP
+
+static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector)
+{
+	struct ht_irq_msg msg;
+	fetch_ht_irq_msg(irq, &msg);
+
+	msg.address_lo &= ~(HT_IRQ_LOW_VECTOR_MASK | HT_IRQ_LOW_DEST_ID_MASK);
+	msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK);
+
+	msg.address_lo |= HT_IRQ_LOW_VECTOR(vector) | HT_IRQ_LOW_DEST_ID(dest);
+	msg.address_hi |= HT_IRQ_HIGH_DEST_ID(dest);
+
+	write_ht_irq_msg(irq, &msg);
+}
+
+static void set_ht_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	unsigned int dest;
+	cpumask_t tmp;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	target_ht_irq(irq, dest, cfg->vector);
+	desc->affinity = 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)
+{
+	struct irq_cfg *cfg;
+	int err;
+	cpumask_t tmp;
+
+	tmp = TARGET_CPUS;
+	err = assign_irq_vector(irq, tmp);
+	if (!err) {
+		struct ht_irq_msg msg;
+		unsigned dest;
+
+		cfg = irq_cfg(irq);
+		cpus_and(tmp, cfg->domain, tmp);
+		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(cfg->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 err;
+}
+#endif /* CONFIG_HT_IRQ */
+
+/* --------------------------------------------------------------------------
+                          ACPI-based IOAPIC Configuration
+   -------------------------------------------------------------------------- */
+
+#ifdef CONFIG_ACPI
+
+#ifdef CONFIG_X86_32
+int __init io_apic_get_unique_id(int ioapic, int apic_id)
+{
+	union IO_APIC_reg_00 reg_00;
+	static physid_mask_t apic_id_map = PHYSID_MASK_NONE;
+	physid_mask_t tmp;
+	unsigned long flags;
+	int i = 0;
+
+	/*
+	 * The P4 platform supports up to 256 APIC IDs on two separate APIC
+	 * buses (one for LAPICs, one for IOAPICs), where predecessors only
+	 * supports up to 16 on one shared APIC bus.
+	 *
+	 * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full
+	 *      advantage of new APIC bus architecture.
+	 */
+
+	if (physids_empty(apic_id_map))
+		apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	reg_00.raw = io_apic_read(ioapic, 0);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	if (apic_id >= get_physical_broadcast()) {
+		printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying "
+			"%d\n", ioapic, apic_id, reg_00.bits.ID);
+		apic_id = reg_00.bits.ID;
+	}
+
+	/*
+	 * Every APIC in a system must have a unique ID or we get lots of nice
+	 * 'stuck on smp_invalidate_needed IPI wait' messages.
+	 */
+	if (check_apicid_used(apic_id_map, apic_id)) {
+
+		for (i = 0; i < get_physical_broadcast(); i++) {
+			if (!check_apicid_used(apic_id_map, i))
+				break;
+		}
+
+		if (i == get_physical_broadcast())
+			panic("Max apic_id exceeded!\n");
+
+		printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
+			"trying %d\n", ioapic, apic_id, i);
+
+		apic_id = i;
+	}
+
+	tmp = apicid_to_cpu_present(apic_id);
+	physids_or(apic_id_map, apic_id_map, tmp);
+
+	if (reg_00.bits.ID != apic_id) {
+		reg_00.bits.ID = apic_id;
+
+		spin_lock_irqsave(&ioapic_lock, flags);
+		io_apic_write(ioapic, 0, reg_00.raw);
+		reg_00.raw = io_apic_read(ioapic, 0);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+
+		/* Sanity check */
+		if (reg_00.bits.ID != apic_id) {
+			printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
+			return -1;
+		}
+	}
+
+	apic_printk(APIC_VERBOSE, KERN_INFO
+			"IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
+
+	return apic_id;
+}
+
+int __init io_apic_get_version(int ioapic)
+{
+	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);
+
+	return reg_01.bits.version;
+}
+#endif
+
+int __init io_apic_get_redir_entries (int ioapic)
+{
+	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);
+
+	return reg_01.bits.entries;
+}
+
+
+int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
+{
+	if (!IO_APIC_IRQ(irq)) {
+		apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
+			ioapic);
+		return -EINVAL;
+	}
+
+	/*
+	 * IRQs < 16 are already in the irq_2_pin[] map
+	 */
+	if (irq >= 16)
+		add_pin_to_irq(irq, ioapic, pin);
+
+	setup_IO_APIC_irq(ioapic, pin, irq, triggering, polarity);
+
+	return 0;
+}
+
+
+int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
+{
+	int i;
+
+	if (skip_ioapic_setup)
+		return -1;
+
+	for (i = 0; i < mp_irq_entries; i++)
+		if (mp_irqs[i].mp_irqtype == mp_INT &&
+		    mp_irqs[i].mp_srcbusirq == bus_irq)
+			break;
+	if (i >= mp_irq_entries)
+		return -1;
+
+	*trigger = irq_trigger(i);
+	*polarity = irq_polarity(i);
+	return 0;
+}
+
+#endif /* CONFIG_ACPI */
+
+/*
+ * This function currently is only a helper for the i386 smp boot process where
+ * we need to reprogram the ioredtbls to cater for the cpus which have come online
+ * so mask in all cases should simply be TARGET_CPUS
+ */
+#ifdef CONFIG_SMP
+void __init setup_ioapic_dest(void)
+{
+	int pin, ioapic, irq, irq_entry;
+	struct irq_cfg *cfg;
+	struct irq_desc *desc;
+
+	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.
+			 */
+			cfg = irq_cfg(irq);
+			if (!cfg->vector)
+				setup_IO_APIC_irq(ioapic, pin, irq,
+						  irq_trigger(irq_entry),
+						  irq_polarity(irq_entry));
+#ifdef CONFIG_INTR_REMAP
+			else if (intr_remapping_enabled) {
+				desc = irq_desc(irq);
+				set_ir_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
+			}
+#endif
+			else {
+				desc = irq_desc(irq);
+				set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
+			}
+		}
+	}
+}
+#endif
+
+#ifdef CONFIG_X86_64
+#define IOAPIC_RESOURCE_NAME_SIZE 11
+
+static struct resource *ioapic_resources;
+
+static struct resource * __init ioapic_setup_resources(void)
+{
+	unsigned long n;
+	struct resource *res;
+	char *mem;
+	int i;
+
+	if (nr_ioapics <= 0)
+		return NULL;
+
+	n = IOAPIC_RESOURCE_NAME_SIZE + sizeof(struct resource);
+	n *= nr_ioapics;
+
+	mem = alloc_bootmem(n);
+	res = (void *)mem;
+
+	if (mem != NULL) {
+		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;
+		}
+	}
+
+	ioapic_resources = res;
+
+	return res;
+}
+#endif
+
+void __init ioapic_init_mappings(void)
+{
+	unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
+	int i;
+#ifdef CONFIG_X86_64
+	struct resource *ioapic_res;
+
+	ioapic_res = ioapic_setup_resources();
+#endif
+	for (i = 0; i < nr_ioapics; i++) {
+		if (smp_found_config) {
+			ioapic_phys = mp_ioapics[i].mp_apicaddr;
+#ifdef CONFIG_X86_32
+                        if (!ioapic_phys) {
+                                printk(KERN_ERR
+                                       "WARNING: bogus zero IO-APIC "
+                                       "address found in MPTABLE, "
+                                       "disabling IO/APIC support!\n");
+                                smp_found_config = 0;
+                                skip_ioapic_setup = 1;
+                                goto fake_ioapic_page;
+                        }
+#endif
+		} else {
+#ifdef CONFIG_X86_32
+fake_ioapic_page:
+#endif
+			ioapic_phys = (unsigned long)
+				alloc_bootmem_pages(PAGE_SIZE);
+			ioapic_phys = __pa(ioapic_phys);
+		}
+		set_fixmap_nocache(idx, ioapic_phys);
+		apic_printk(APIC_VERBOSE,
+			    "mapped IOAPIC to %08lx (%08lx)\n",
+			    __fix_to_virt(idx), ioapic_phys);
+		idx++;
+
+#ifdef CONFIG_X86_64
+		if (ioapic_res != NULL) {
+			ioapic_res->start = ioapic_phys;
+			ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
+			ioapic_res++;
+		}
+#endif
+	}
+}
+
+#ifdef CONFIG_X86_64
+static int __init ioapic_insert_resources(void)
+{
+	int i;
+	struct resource *r = ioapic_resources;
+
+	if (!r) {
+		printk(KERN_ERR
+		       "IO APIC resources could be not be allocated.\n");
+		return -1;
+	}
+
+	for (i = 0; i < nr_ioapics; i++) {
+		insert_resource(&iomem_resource, r);
+		r++;
+	}
+
+	return 0;
+}
+
+/* Insert the IO APIC resources after PCI initialization has occured to handle
+ * IO APICS that are mapped in on a BAR in PCI space. */
+late_initcall(ioapic_insert_resources);
+#endif
Index: linux-2.6/arch/x86/kernel/io_apic_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_32.c
+++ /dev/null
@@ -1,3220 +0,0 @@
-/*
- *	Intel IO-APIC support for multi-Pentium hosts.
- *
- *	Copyright (C) 1997, 1998, 1999, 2000 Ingo Molnar, Hajnalka Szabo
- *
- *	Many thanks to Stig Venaas for trying out countless experimental
- *	patches and reporting/debugging problems patiently!
- *
- *	(c) 1999, Multiple IO-APIC support, developed by
- *	Ken-ichi Yaku <yaku@css1.kbnes.nec.co.jp> and
- *      Hidemi Kishimoto <kisimoto@css1.kbnes.nec.co.jp>,
- *	further tested and cleaned up by Zach Brown <zab@redhat.com>
- *	and Ingo Molnar <mingo@redhat.com>
- *
- *	Fixes
- *	Maciej W. Rozycki	:	Bits for genuine 82489DX APICs;
- *					thanks to Eric Gilmore
- *					and Rolf G. Tews
- *					for testing these extensively
- *	Paul Diefenbaugh	:	Added full ACPI support
- */
-
-#include <linux/mm.h>
-#include <linux/interrupt.h>
-#include <linux/init.h>
-#include <linux/delay.h>
-#include <linux/sched.h>
-#include <linux/pci.h>
-#include <linux/mc146818rtc.h>
-#include <linux/compiler.h>
-#include <linux/acpi.h>
-#include <linux/module.h>
-#include <linux/sysdev.h>
-#include <linux/msi.h>
-#include <linux/htirq.h>
-#include <linux/freezer.h>
-#include <linux/kthread.h>
-#include <linux/jiffies.h>      /* time_after() */
-#ifdef CONFIG_ACPI
-#include <acpi/acpi_bus.h>
-#endif
-#include <linux/bootmem.h>
-#include <linux/dmar.h>
-
-#include <asm/idle.h>
-#include <asm/io.h>
-#include <asm/smp.h>
-#include <asm/desc.h>
-#include <asm/proto.h>
-#include <asm/acpi.h>
-#include <asm/dma.h>
-#include <asm/timer.h>
-#include <asm/i8259.h>
-#include <asm/nmi.h>
-#include <asm/msidef.h>
-#include <asm/hypertransport.h>
-#include <asm/setup.h>
-#include <asm/irq_remapping.h>
-
-#include <mach_ipi.h>
-#include <mach_apic.h>
-#include <mach_apicdef.h>
-
-#define __apicdebuginit(type) static type __init
-
-/*
- *	Is the SiS APIC rmw bug present ?
- *	-1 = don't know, 0 = no, 1 = yes
- */
-int sis_apic_bug = -1;
-
-static DEFINE_SPINLOCK(ioapic_lock);
-static DEFINE_SPINLOCK(vector_lock);
-
-int first_free_entry;
-/*
- * Rough estimation of how many shared IRQs there are, can
- * be changed anytime.
- */
-int pin_map_size;
-
-/*
- * # of IRQ routing registers
- */
-int nr_ioapic_registers[MAX_IO_APICS];
-
-/* I/O APIC entries */
-struct mp_config_ioapic mp_ioapics[MAX_IO_APICS];
-int nr_ioapics;
-
-/* MP IRQ source entries */
-struct mp_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
-
-/* # of MP IRQ source entries */
-int mp_irq_entries;
-
-#if defined (CONFIG_MCA) || defined (CONFIG_EISA)
-int mp_bus_id_to_type[MAX_MP_BUSSES];
-#endif
-
-DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES);
-
-int skip_ioapic_setup;
-
-static int __init parse_noapic(char *arg)
-{
-	/* disable IO-APIC */
-	disable_ioapic_setup();
-	return 0;
-}
-early_param("noapic", parse_noapic);
-
-struct irq_cfg;
-struct irq_pin_list;
-struct irq_cfg {
-	unsigned int irq;
-	struct irq_cfg *next;
-	struct irq_pin_list *irq_2_pin;
-	cpumask_t domain;
-	cpumask_t old_domain;
-	unsigned move_cleanup_count;
-	u8 vector;
-	u8 move_in_progress : 1;
-};
-
-/* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
-static struct irq_cfg irq_cfg_legacy[] __initdata = {
-	[0]  = { .irq =  0, .domain = CPU_MASK_ALL, .vector = IRQ0_VECTOR,  },
-	[1]  = { .irq =  1, .domain = CPU_MASK_ALL, .vector = IRQ1_VECTOR,  },
-	[2]  = { .irq =  2, .domain = CPU_MASK_ALL, .vector = IRQ2_VECTOR,  },
-	[3]  = { .irq =  3, .domain = CPU_MASK_ALL, .vector = IRQ3_VECTOR,  },
-	[4]  = { .irq =  4, .domain = CPU_MASK_ALL, .vector = IRQ4_VECTOR,  },
-	[5]  = { .irq =  5, .domain = CPU_MASK_ALL, .vector = IRQ5_VECTOR,  },
-	[6]  = { .irq =  6, .domain = CPU_MASK_ALL, .vector = IRQ6_VECTOR,  },
-	[7]  = { .irq =  7, .domain = CPU_MASK_ALL, .vector = IRQ7_VECTOR,  },
-	[8]  = { .irq =  8, .domain = CPU_MASK_ALL, .vector = IRQ8_VECTOR,  },
-	[9]  = { .irq =  9, .domain = CPU_MASK_ALL, .vector = IRQ9_VECTOR,  },
-	[10] = { .irq = 10, .domain = CPU_MASK_ALL, .vector = IRQ10_VECTOR, },
-	[11] = { .irq = 11, .domain = CPU_MASK_ALL, .vector = IRQ11_VECTOR, },
-	[12] = { .irq = 12, .domain = CPU_MASK_ALL, .vector = IRQ12_VECTOR, },
-	[13] = { .irq = 13, .domain = CPU_MASK_ALL, .vector = IRQ13_VECTOR, },
-	[14] = { .irq = 14, .domain = CPU_MASK_ALL, .vector = IRQ14_VECTOR, },
-	[15] = { .irq = 15, .domain = CPU_MASK_ALL, .vector = IRQ15_VECTOR, },
-};
-
-static struct irq_cfg irq_cfg_init = { .irq =  -1U, };
-/* need to be biger than size of irq_cfg_legacy */
-static int nr_irq_cfg = 32;
-
-static int __init parse_nr_irq_cfg(char *arg)
-{
-	if (arg) {
-		nr_irq_cfg = simple_strtoul(arg, NULL, 0);
-		if (nr_irq_cfg < 32)
-			nr_irq_cfg = 32;
-	}
-	return 0;
-}
-
-early_param("nr_irq_cfg", parse_nr_irq_cfg);
-
-static void init_one_irq_cfg(struct irq_cfg *cfg)
-{
-	memcpy(cfg, &irq_cfg_init, sizeof(struct irq_cfg));
-}
-
-static struct irq_cfg *irq_cfgx;
-static struct irq_cfg *irq_cfgx_free;
-static void __init init_work(void *data)
-{
-	struct dyn_array *da = data;
-	struct irq_cfg *cfg;
-	int legacy_count;
-	int i;
-
-	cfg = *da->name;
-
-	memcpy(cfg, irq_cfg_legacy, sizeof(irq_cfg_legacy));
-
-	legacy_count = sizeof(irq_cfg_legacy)/sizeof(irq_cfg_legacy[0]);
-	for (i = legacy_count; i < *da->nr; i++)
-		init_one_irq_cfg(&cfg[i]);
-
-	for (i = 1; i < *da->nr; i++)
-		cfg[i-1].next = &cfg[i];
-
-	irq_cfgx_free = &irq_cfgx[legacy_count];
-	irq_cfgx[legacy_count - 1].next = NULL;
-}
-
-#define for_each_irq_cfg(cfg)           \
-	for (cfg = irq_cfgx; cfg; cfg = cfg->next)
-
-DEFINE_DYN_ARRAY(irq_cfgx, sizeof(struct irq_cfg), nr_irq_cfg, PAGE_SIZE, init_work);
-
-static struct irq_cfg *irq_cfg(unsigned int irq)
-{
-	struct irq_cfg *cfg;
-
-	cfg = irq_cfgx;
-	while (cfg) {
-		if (cfg->irq == irq)
-			return cfg;
-
-		cfg = cfg->next;
-	}
-
-	return NULL;
-}
-
-static struct irq_cfg *irq_cfg_with_new(unsigned int irq)
-{
-	struct irq_cfg *cfg, *cfg_pri;
-	int i;
-	int count = 0;
-
-	cfg_pri = cfg = irq_cfgx;
-	while (cfg) {
-		if (cfg->irq == irq)
-			return cfg;
-
-		cfg_pri = cfg;
-		cfg = cfg->next;
-		count++;
-	}
-
-	if (!irq_cfgx_free) {
-		unsigned long phys;
-		unsigned long total_bytes;
-		/*
-		 *  we run out of pre-allocate ones, allocate more
-		 */
-		printk(KERN_DEBUG "try to get more irq_cfg %d\n", nr_irq_cfg);
-
-		total_bytes = sizeof(struct irq_cfg) * nr_irq_cfg;
-		if (after_bootmem)
-			cfg = kzalloc(total_bytes, GFP_ATOMIC);
-		else
-			cfg = __alloc_bootmem_nopanic(total_bytes, PAGE_SIZE, 0);
-
-		if (!cfg)
-			panic("please boot with nr_irq_cfg= %d\n", count * 2);
-
-		phys = __pa(cfg);
-		printk(KERN_DEBUG "irq_irq ==> [%#lx - %#lx]\n", phys, phys + total_bytes);
-
-		for (i = 0; i < nr_irq_cfg; i++)
-			init_one_irq_cfg(&cfg[i]);
-
-		for (i = 1; i < nr_irq_cfg; i++)
-			cfg[i-1].next = &cfg[i];
-
-		irq_cfgx_free = cfg;
-	}
-
-	cfg = irq_cfgx_free;
-	irq_cfgx_free = irq_cfgx_free->next;
-	cfg->next = NULL;
-	if (cfg_pri)
-		cfg_pri->next = cfg;
-	else
-		irq_cfgx = cfg;
-	cfg->irq = irq;
-	printk(KERN_DEBUG "found new irq_cfg for irq %d\n", cfg->irq);
-
-#ifdef CONFIG_HAVE_SPARSE_IRQ_DEBUG
-	{
-		/* dump the results */
-		struct irq_cfg *cfg;
-		unsigned long phys;
-		unsigned long bytes = sizeof(struct irq_cfg);
-
-		printk(KERN_DEBUG "=========================== %d\n", irq);
-		printk(KERN_DEBUG "irq_cfg dump after get that for %d\n", irq);
-		for_each_irq_cfg(cfg) {
-			phys = __pa(cfg);
-			printk(KERN_DEBUG "irq_cfg %d ==> [%#lx - %#lx]\n", cfg->irq, phys, phys + bytes);
-		}
-		printk(KERN_DEBUG "===========================\n");
-	}
-#endif
-	return cfg;
-}
-
-/*
- * 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 *irq_2_pin_head;
-/* fill one page ? */
-static int nr_irq_2_pin = 0x100;
-static struct irq_pin_list *irq_2_pin_ptr;
-static void __init irq_2_pin_init_work(void *data)
-{
-	struct dyn_array *da = data;
-	struct irq_pin_list *pin;
-	int i;
-
-	pin = *da->name;
-
-	for (i = 1; i < *da->nr; i++)
-		pin[i-1].next = &pin[i];
-
-	irq_2_pin_ptr = &pin[0];
-}
-DEFINE_DYN_ARRAY(irq_2_pin_head, sizeof(struct irq_pin_list), nr_irq_2_pin, PAGE_SIZE, irq_2_pin_init_work);
-
-static struct irq_pin_list *get_one_free_irq_2_pin(void)
-{
-	struct irq_pin_list *pin;
-	int i;
-
-	pin = irq_2_pin_ptr;
-
-	if (pin) {
-		irq_2_pin_ptr = pin->next;
-		pin->next = NULL;
-		return pin;
-	}
-
-	/*
-	 *  we run out of pre-allocate ones, allocate more
-	 */
-	printk(KERN_DEBUG "try to get more irq_2_pin %d\n", nr_irq_2_pin);
-
-	if (after_bootmem)
-		pin = kzalloc(sizeof(struct irq_pin_list)*nr_irq_2_pin,
-				 GFP_ATOMIC);
-	else
-		pin = __alloc_bootmem_nopanic(sizeof(struct irq_pin_list) *
-				nr_irq_2_pin, PAGE_SIZE, 0);
-
-	if (!pin)
-		panic("can not get more irq_2_pin\n");
-
-	for (i = 1; i < nr_irq_2_pin; i++)
-		pin[i-1].next = &pin[i];
-
-	irq_2_pin_ptr = pin->next;
-	pin->next = NULL;
-
-	return pin;
-}
-
-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].mp_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.
- *
- * Older SiS APIC requires we rewrite the index register
- */
-static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
-{
-	volatile struct io_apic __iomem *io_apic = io_apic_base(apic);
-	if (sis_apic_bug)
-		writel(reg, &io_apic->index);
-	writel(value, &io_apic->data);
-}
-
-#ifdef CONFIG_X86_64
-static bool io_apic_level_ack_pending(unsigned int irq)
-{
-	struct irq_pin_list *entry;
-	unsigned long flags;
-	struct irq_cfg *cfg = irq_cfg(irq);
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		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? */
-		if (reg & IO_APIC_REDIR_REMOTE_IRR) {
-			spin_unlock_irqrestore(&ioapic_lock, flags);
-			return true;
-		}
-		if (!entry->next)
-			break;
-		entry = entry->next;
-	}
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	return false;
-}
-#endif
-
-union entry_union {
-	struct { u32 w1, w2; };
-	struct IO_APIC_route_entry entry;
-};
-
-static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin)
-{
-	union entry_union eu;
-	unsigned long flags;
-	spin_lock_irqsave(&ioapic_lock, flags);
-	eu.w1 = io_apic_read(apic, 0x10 + 2 * pin);
-	eu.w2 = io_apic_read(apic, 0x11 + 2 * pin);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-	return eu.entry;
-}
-
-/*
- * When we write a new IO APIC routing entry, we need to write the high
- * word first! If the mask bit in the low word is clear, we will enable
- * the interrupt, and we need to make sure the entry is fully populated
- * before that happens.
- */
-static void
-__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
-{
-	union entry_union eu;
-	eu.entry = e;
-	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
-	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
-}
-
-static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
-{
-	unsigned long flags;
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__ioapic_write_entry(apic, pin, e);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-/*
- * When we mask an IO APIC routing entry, we need to write the low
- * word first, in order to set the mask bit before we change the
- * high bits!
- */
-static void ioapic_mask_entry(int apic, int pin)
-{
-	unsigned long flags;
-	union entry_union eu = { .entry.mask = 1 };
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
-	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-#ifdef CONFIG_SMP
-static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, u8 vector)
-{
-	int apic, pin;
-	struct irq_cfg *cfg;
-	struct irq_pin_list *entry;
-
-	cfg = irq_cfg(irq);
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		unsigned int reg;
-
-		if (!entry)
-			break;
-
-		apic = entry->apic;
-		pin = entry->pin;
-		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, cpumask_t mask);
-
-static void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	unsigned long flags;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	cfg = irq_cfg(irq);
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-	/*
-	 * Only the high 8 bits are valid.
-	 */
-	dest = SET_APIC_LOGICAL_ID(dest);
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__target_IO_APIC_irq(irq, dest, cfg->vector);
-	desc->affinity = mask;
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-#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(unsigned int irq, int apic, int pin)
-{
-	struct irq_cfg *cfg;
-	struct irq_pin_list *entry;
-
-	/* first time to refer irq_cfg, so with new */
-	cfg = irq_cfg_with_new(irq);
-	entry = cfg->irq_2_pin;
-	if (!entry) {
-		entry = get_one_free_irq_2_pin();
-		cfg->irq_2_pin = entry;
-		entry->apic = apic;
-		entry->pin = pin;
-		printk(KERN_DEBUG " 0 add_pin_to_irq: irq %d --> apic %d pin %d\n", irq, apic, pin);
-		return;
-	}
-
-	while (entry->next) {
-		/* not again, please */
-		if (entry->apic == apic && entry->pin == pin)
-			return;
-
-		entry = entry->next;
-	}
-
-	entry->next = get_one_free_irq_2_pin();
-	entry = entry->next;
-	entry->apic = apic;
-	entry->pin = pin;
-	printk(KERN_DEBUG " x add_pin_to_irq: irq %d --> apic %d pin %d\n", irq, apic, pin);
-}
-
-/*
- * Reroute an IRQ to a different pin.
- */
-static void __init replace_pin_at_irq(unsigned int irq,
-				      int oldapic, int oldpin,
-				      int newapic, int newpin)
-{
-	struct irq_cfg *cfg = irq_cfg(irq);
-	struct irq_pin_list *entry = cfg->irq_2_pin;
-	int replaced = 0;
-
-	while (entry) {
-		if (entry->apic == oldapic && entry->pin == oldpin) {
-			entry->apic = newapic;
-			entry->pin = newpin;
-			replaced = 1;
-			/* every one is different, right? */
-			break;
-		}
-		entry = entry->next;
-	}
-
-	/* why? call replace before add? */
-	if (!replaced)
-		add_pin_to_irq(irq, newapic, newpin);
-}
-
-#ifdef CONFIG_X86_64
-/*
- * Synchronize the IO-APIC and the CPU by doing
- * a dummy read from the IO-APIC
- */
-static inline void io_apic_sync(unsigned int apic)
-{
-	struct io_apic __iomem *io_apic = io_apic_base(apic);
-	readl(&io_apic->data);
-}
-
-#define __DO_ACTION(R, ACTION, FINAL)					\
-									\
-{									\
-	int pin;							\
-	struct irq_cfg *cfg;						\
-	struct irq_pin_list *entry;					\
-									\
-	cfg = irq_cfg(irq);						\
-	entry = cfg->irq_2_pin;						\
-	for (;;) {							\
-		unsigned int reg;					\
-		if (!entry)						\
-			break;						\
-		pin = entry->pin;					\
-		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
-		reg ACTION;						\
-		io_apic_modify(entry->apic, 0x10 + R + pin*2, reg);	\
-		FINAL;							\
-		if (!entry->next)					\
-			break;						\
-		entry = entry->next;					\
-	}								\
-}
-
-#define DO_ACTION(name,R,ACTION, FINAL)					\
-									\
-	static void name##_IO_APIC_irq (unsigned int irq)		\
-	__DO_ACTION(R, ACTION, FINAL)
-
-/* mask = 1 */
-DO_ACTION(__mask,	0, |= IO_APIC_REDIR_MASKED, io_apic_sync(entry->apic))
-
-/* mask = 0 */
-DO_ACTION(__unmask,	0, &= ~IO_APIC_REDIR_MASKED, )
-
-#else
-
-static void __modify_IO_APIC_irq(unsigned int irq, unsigned long enable, unsigned long disable)
-{
-	struct irq_cfg *cfg;
-	struct irq_pin_list *entry;
-	unsigned int pin, reg;
-
-	cfg = irq_cfg(irq);
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		if (!entry)
-			break;
-		pin = entry->pin;
-		reg = io_apic_read(entry->apic, 0x10 + pin*2);
-		reg &= ~disable;
-		reg |= enable;
-		io_apic_modify(entry->apic, 0x10 + pin*2, reg);
-		if (!entry->next)
-			break;
-		entry = entry->next;
-	}
-}
-
-/* mask = 1 */
-static void __mask_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED, 0);
-}
-
-/* mask = 0 */
-static void __unmask_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, 0, IO_APIC_REDIR_MASKED);
-}
-
-/* mask = 1, trigger = 0 */
-static void __mask_and_edge_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED,
-				IO_APIC_REDIR_LEVEL_TRIGGER);
-}
-
-/* mask = 0, trigger = 1 */
-static void __unmask_and_level_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_LEVEL_TRIGGER,
-				IO_APIC_REDIR_MASKED);
-}
-
-#endif
-
-static void mask_IO_APIC_irq(unsigned int irq)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__mask_IO_APIC_irq(irq);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-static void unmask_IO_APIC_irq(unsigned int irq)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__unmask_IO_APIC_irq(irq);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
-{
-	struct IO_APIC_route_entry entry;
-
-	/* Check delivery_mode to be sure we're not clearing an SMI pin */
-	entry = ioapic_read_entry(apic, pin);
-	if (entry.delivery_mode == dest_SMI)
-		return;
-
-	/*
-	 * Disable it in the IO-APIC irq-routing table:
-	 */
-	ioapic_mask_entry(apic, pin);
-}
-
-static void clear_IO_APIC(void)
-{
-	int apic, pin;
-
-	for (apic = 0; apic < nr_ioapics; apic++)
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
-			clear_IO_APIC_pin(apic, pin);
-}
-
-#ifndef CONFIG_SMP
-void send_IPI_self(int vector)
-{
-	unsigned int cfg;
-
-	/*
-	 * Wait for idle.
-	 */
-	apic_wait_icr_idle();
-	cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
-	/*
-	 * Send the IPI. The write to APIC_ICR fires this off.
-	 */
-	apic_write(APIC_ICR, cfg);
-}
-#endif /* !CONFIG_SMP */
-
-
-/*
- * 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;
-
-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, KERN_INFO
-			"PIRQ redirection, working around broken MP-BIOS.\n");
-	max = MAX_PIRQS;
-	if (ints[0] < MAX_PIRQS)
-		max = ints[0];
-
-	for (i = 0; i < max; i++) {
-		apic_printk(APIC_VERBOSE, KERN_DEBUG
-				"... PIRQ%d -> IRQ %d\n", i, ints[i+1]);
-		/*
-		 * PIRQs are mapped upside down, usually.
-		 */
-		pirq_entries[MAX_PIRQS-i-1] = ints[i+1];
-	}
-	return 1;
-}
-
-__setup("pirq=", ioapic_pirq_setup);
-
-/*
- * Find the IRQ entry number of a certain pin.
- */
-static int find_irq_entry(int apic, int pin, int type)
-{
-	int i;
-
-	for (i = 0; i < mp_irq_entries; i++)
-		if (mp_irqs[i].mp_irqtype == type &&
-		    (mp_irqs[i].mp_dstapic == mp_ioapics[apic].mp_apicid ||
-		     mp_irqs[i].mp_dstapic == MP_APIC_ALL) &&
-		    mp_irqs[i].mp_dstirq == pin)
-			return i;
-
-	return -1;
-}
-
-/*
- * Find the pin to which IRQ[irq] (ISA) is connected
- */
-static int __init find_isa_irq_pin(int irq, int type)
-{
-	int i;
-
-	for (i = 0; i < mp_irq_entries; i++) {
-		int lbus = mp_irqs[i].mp_srcbus;
-
-		if (test_bit(lbus, mp_bus_not_pci) &&
-		    (mp_irqs[i].mp_irqtype == type) &&
-		    (mp_irqs[i].mp_srcbusirq == irq))
-
-			return mp_irqs[i].mp_dstirq;
-	}
-	return -1;
-}
-
-static int __init find_isa_irq_apic(int irq, int type)
-{
-	int i;
-
-	for (i = 0; i < mp_irq_entries; i++) {
-		int lbus = mp_irqs[i].mp_srcbus;
-
-		if (test_bit(lbus, mp_bus_not_pci) &&
-		    (mp_irqs[i].mp_irqtype == type) &&
-		    (mp_irqs[i].mp_srcbusirq == irq))
-			break;
-	}
-	if (i < mp_irq_entries) {
-		int apic;
-		for (apic = 0; apic < nr_ioapics; apic++) {
-			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic)
-				return apic;
-		}
-	}
-
-	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)) {
-		printk(KERN_WARNING "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
-		return -1;
-	}
-	for (i = 0; i < mp_irq_entries; i++) {
-		int lbus = mp_irqs[i].mp_srcbus;
-
-		for (apic = 0; apic < nr_ioapics; apic++)
-			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic ||
-			    mp_irqs[i].mp_dstapic == MP_APIC_ALL)
-				break;
-
-		if (!test_bit(lbus, mp_bus_not_pci) &&
-		    !mp_irqs[i].mp_irqtype &&
-		    (bus == lbus) &&
-		    (slot == ((mp_irqs[i].mp_srcbusirq >> 2) & 0x1f))) {
-			int irq = pin_2_irq(i, apic, mp_irqs[i].mp_dstirq);
-
-			if (!(apic || IO_APIC_IRQ(irq)))
-				continue;
-
-			if (pin == (mp_irqs[i].mp_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
- */
-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, KERN_INFO
-			"Broken MPtable reports ISA irq %d\n", irq);
-	return 0;
-}
-#endif
-
-/* ISA interrupts are always polarity zero edge triggered,
- * when listed as conforming in the MP table. */
-
-#define default_ISA_trigger(idx)	(0)
-#define default_ISA_polarity(idx)	(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].mp_srcbusirq))
-#define default_EISA_polarity(idx)	default_ISA_polarity(idx)
-
-/* PCI interrupts are always polarity one level triggered,
- * when listed as conforming in the MP table. */
-
-#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)	default_ISA_polarity(idx)
-
-static int MPBIOS_polarity(int idx)
-{
-	int bus = mp_irqs[idx].mp_srcbus;
-	int polarity;
-
-	/*
-	 * Determine IRQ line polarity (high active or low active):
-	 */
-	switch (mp_irqs[idx].mp_irqflag & 3) {
-	case 0: /* conforms, ie. bus-type dependent polarity */
-	{
-		polarity = test_bit(bus, mp_bus_not_pci)?
-			default_ISA_polarity(idx):
-			default_PCI_polarity(idx);
-		break;
-	}
-	case 1: /* high active */
-	{
-		polarity = 0;
-		break;
-	}
-	case 2: /* reserved */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		polarity = 1;
-		break;
-	}
-	case 3: /* low active */
-	{
-		polarity = 1;
-		break;
-	}
-	default: /* invalid */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		polarity = 1;
-		break;
-	}
-	}
-	return polarity;
-}
-
-static int MPBIOS_trigger(int idx)
-{
-	int bus = mp_irqs[idx].mp_srcbus;
-	int trigger;
-
-	/*
-	 * Determine IRQ trigger mode (edge or level sensitive):
-	 */
-	switch ((mp_irqs[idx].mp_irqflag>>2) & 3) {
-	case 0: /* conforms, ie. bus-type dependent */
-	{
-		trigger = test_bit(bus, mp_bus_not_pci)?
-				default_ISA_trigger(idx):
-				default_PCI_trigger(idx);
-#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
-		switch (mp_bus_id_to_type[bus]) {
-		case MP_BUS_ISA: /* ISA pin */
-		{
-			/* set before the switch */
-			break;
-		}
-		case MP_BUS_EISA: /* EISA pin */
-		{
-			trigger = default_EISA_trigger(idx);
-			break;
-		}
-		case MP_BUS_PCI: /* PCI pin */
-		{
-			/* set before the switch */
-			break;
-		}
-		case MP_BUS_MCA: /* MCA pin */
-		{
-			trigger = default_MCA_trigger(idx);
-			break;
-		}
-		default:
-		{
-			printk(KERN_WARNING "broken BIOS!!\n");
-			trigger = 1;
-			break;
-		}
-	}
-#endif
-		break;
-	}
-	case 1: /* edge */
-	{
-		trigger = 0;
-		break;
-	}
-	case 2: /* reserved */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		trigger = 1;
-		break;
-	}
-	case 3: /* level */
-	{
-		trigger = 1;
-		break;
-	}
-	default: /* invalid */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		trigger = 0;
-		break;
-	}
-	}
-	return trigger;
-}
-
-static inline int irq_polarity(int idx)
-{
-	return MPBIOS_polarity(idx);
-}
-
-static inline int irq_trigger(int idx)
-{
-	return MPBIOS_trigger(idx);
-}
-
-int (*ioapic_renumber_irq)(int ioapic, int irq);
-static int pin_2_irq(int idx, int apic, int pin)
-{
-	int irq, i;
-	int bus = mp_irqs[idx].mp_srcbus;
-
-	/*
-	 * Debugging check, we are in big trouble if this message pops up!
-	 */
-	if (mp_irqs[idx].mp_dstirq != pin)
-		printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
-
-	if (test_bit(bus, mp_bus_not_pci))
-		irq = mp_irqs[idx].mp_srcbusirq;
-	else {
-		/*
-		 * PCI IRQs are mapped in order
-		 */
-		i = irq = 0;
-		while (i < apic)
-			irq += nr_ioapic_registers[i++];
-		irq += pin;
-
-		/*
-		 * For MPS mode, so far only needed by ES7000 platform
-		 */
-		if (ioapic_renumber_irq)
-			irq = ioapic_renumber_irq(apic, irq);
-	}
-
-	/*
-	 * PCI IRQ command line redirection. Yes, limits are hardcoded.
-	 */
-	if ((pin >= 16) && (pin <= 23)) {
-		if (pirq_entries[pin-16] != -1) {
-			if (!pirq_entries[pin-16]) {
-				apic_printk(APIC_VERBOSE, KERN_DEBUG
-						"disabling PIRQ%d\n", pin-16);
-			} else {
-				irq = pirq_entries[pin-16];
-				apic_printk(APIC_VERBOSE, KERN_DEBUG
-						"using PIRQ%d -> IRQ %d\n",
-						pin-16, irq);
-			}
-		}
-	}
-	return irq;
-}
-
-void lock_vector_lock(void)
-{
-	/* Used to the online set of cpus does not change
-	 * during assign_irq_vector.
-	 */
-	spin_lock(&vector_lock);
-}
-
-void unlock_vector_lock(void)
-{
-	spin_unlock(&vector_lock);
-}
-
-static int __assign_irq_vector(int irq, cpumask_t mask)
-{
-	/*
-	 * NOTE! The local APIC isn't very good at handling
-	 * multiple interrupts at the same interrupt level.
-	 * As the interrupt level is determined by taking the
-	 * vector number and shifting that right by 4, we
-	 * want to spread these out a bit so that they don't
-	 * all fall in the same interrupt level.
-	 *
-	 * Also, we've got to be careful not to trash gate
-	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
-	 */
-        static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0;
-        unsigned int old_vector;
-        int cpu;
-        struct irq_cfg *cfg;
-
-        cfg = irq_cfg(irq);
-
-        /* Only try and allocate irqs on cpus that are present */
-        cpus_and(mask, mask, cpu_online_map);
-
-        if ((cfg->move_in_progress) || cfg->move_cleanup_count)
-                return -EBUSY;
-
-        old_vector = cfg->vector;
-        if (old_vector) {
-                cpumask_t tmp;
-                cpus_and(tmp, cfg->domain, mask);
-                if (!cpus_empty(tmp))
-                        return 0;
-        }
-
-        for_each_cpu_mask_nr(cpu, mask) {
-                cpumask_t domain, new_mask;
-                int new_cpu;
-                int vector, offset;
-
-                domain = vector_allocation_domain(cpu);
-                cpus_and(new_mask, domain, cpu_online_map);
-
-                vector = current_vector;
-                offset = current_offset;
-next:
-                vector += 8;
-                if (vector >= first_system_vector) {
-                        /* If we run out of vectors on large boxen, must share them. */
-                        offset = (offset + 1) % 8;
-                        vector = FIRST_DEVICE_VECTOR + offset;
-                }
-                if (unlikely(current_vector == vector))
-                        continue;
-#ifdef CONFIG_X86_64
-                if (vector == IA32_SYSCALL_VECTOR)
-                        goto next;
-#else
-                if (vector == SYSCALL_VECTOR)
-                        goto next;
-#endif
-                for_each_cpu_mask_nr(new_cpu, new_mask)
-                        if (per_cpu(vector_irq, new_cpu)[vector] != -1)
-                                goto next;
-                /* Found one! */
-                current_vector = vector;
-                current_offset = offset;
-                if (old_vector) {
-                        cfg->move_in_progress = 1;
-                        cfg->old_domain = cfg->domain;
-                }
-		printk(KERN_DEBUG "assign_irq_vector: irq %d vector %#x cpu ", irq, vector);
-		for_each_cpu_mask_nr(new_cpu, new_mask) {
-			per_cpu(vector_irq, new_cpu)[vector] = irq;
-			printk(KERN_CONT " %d ", new_cpu);
-		}
-		printk(KERN_CONT "\n");
-                cfg->vector = vector;
-                cfg->domain = domain;
-                return 0;
-        }
-        return -ENOSPC;
-}
-
-static int assign_irq_vector(int irq, cpumask_t mask)
-{
-	int err;
-	unsigned long flags;
-
-	spin_lock_irqsave(&vector_lock, flags);
-	err = __assign_irq_vector(irq, mask);
-	spin_unlock_irqrestore(&vector_lock, flags);
-
-	return err;
-}
-
-static void __clear_irq_vector(int irq)
-{
-	struct irq_cfg *cfg;
-	cpumask_t mask;
-	int cpu, vector;
-
-	cfg = irq_cfg(irq);
-	BUG_ON(!cfg->vector);
-
-	vector = cfg->vector;
-	cpus_and(mask, cfg->domain, cpu_online_map);
-	for_each_cpu_mask_nr(cpu, mask)
-		per_cpu(vector_irq, cpu)[vector] = -1;
-
-	cfg->vector = 0;
-	cpus_clear(cfg->domain);
-}
-
-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;
-	struct irq_cfg *cfg;
-
-	/* Mark the inuse vectors */
-	for_each_irq_cfg(cfg) {
-		if (!cpu_isset(cpu, cfg->domain))
-			continue;
-		vector = cfg->vector;
-		irq = cfg->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;
-
-		cfg = irq_cfg(irq);
-		if (!cpu_isset(cpu, cfg->domain))
-			per_cpu(vector_irq, cpu)[vector] = -1;
-        }
-}
-
-static struct irq_chip ioapic_chip;
-
-#define IOAPIC_AUTO	-1
-#define IOAPIC_EDGE	0
-#define IOAPIC_LEVEL	1
-
-#ifdef CONFIG_X86_32
-static inline int IO_APIC_irq_trigger(int irq)
-{
-	int apic, idx, pin;
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-			idx = find_irq_entry(apic, pin, mp_INT);
-			if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
-				return irq_trigger(idx);
-		}
-	}
-	/*
-	 * nonexistent IRQs are edge default
-	 */
-	return 0;
-}
-#else
-static inline int IO_APIC_irq_trigger(int irq)
-{
-        return 1;
-}
-#endif
-
-static void ioapic_register_intr(int irq, unsigned long trigger)
-{
-	struct irq_desc *desc;
-
-	/* first time to use this irq_desc */
-	if (irq < 16)
-		desc = irq_desc(irq);
-	else
-		desc = irq_desc_with_new(irq);
-
-	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-	    trigger == IOAPIC_LEVEL)
-		desc->status |= IRQ_LEVEL;
-	else
-		desc->status &= ~IRQ_LEVEL;
-
-	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-	    trigger == IOAPIC_LEVEL)
-		set_irq_chip_and_handler_name(irq, &ioapic_chip,
-					 handle_fasteoi_irq, "fasteoi");
-	else
-		set_irq_chip_and_handler_name(irq, &ioapic_chip,
-					 handle_edge_irq, "edge");
-}
-
-static int setup_ioapic_entry(int apic, int irq,
-			      struct IO_APIC_route_entry *entry,
-			      unsigned int destination, int trigger,
-			      int polarity, int vector)
-{
-	/*
-	 * 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->dest = destination;
-
-	entry->mask = 0;                                /* enable IRQ */
-	entry->trigger = trigger;
-	entry->polarity = polarity;
-	entry->vector = vector;
-
-	/* Mask level triggered irqs.
-	 * Use IRQ_DELAYED_DISABLE for edge triggered irqs.
-	 */
-	if (trigger)
-		entry->mask = 1;
-
-	return 0;
-}
-
-static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq,
-                              int trigger, int polarity)
-{
-	struct irq_cfg *cfg;
-	struct IO_APIC_route_entry entry;
-	cpumask_t mask;
-
-	if (!IO_APIC_IRQ(irq))
-		return;
-
-	cfg = irq_cfg(irq);
-
-	mask = TARGET_CPUS;
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cpus_and(mask, cfg->domain, mask);
-
-	apic_printk(APIC_VERBOSE,KERN_DEBUG
-		    "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> "
-		    "IRQ %d Mode:%i Active:%i)\n",
-		    apic, mp_ioapics[apic].mp_apicid, pin, cfg->vector,
-		    irq, trigger, polarity);
-
-
-	if (setup_ioapic_entry(mp_ioapics[apic].mp_apicid, irq, &entry,
-			       cpu_mask_to_apicid(mask), trigger, polarity,
-			       cfg->vector)) {
-		printk("Failed to setup ioapic entry for ioapic  %d, pin %d\n",
-		       mp_ioapics[apic].mp_apicid, pin);
-		__clear_irq_vector(irq);
-		return;
-	}
-
-	ioapic_register_intr(irq, trigger);
-	if (irq < 16)
-		disable_8259A_irq(irq);
-
-	ioapic_write_entry(apic, pin, entry);
-}
-
-static void __init setup_IO_APIC_irqs(void)
-{
-	int apic, pin, idx, irq, first_notcon = 1;
-
-	apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-	for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-
-		idx = find_irq_entry(apic,pin,mp_INT);
-		if (idx == -1) {
-			if (first_notcon) {
-				apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mp_apicid, pin);
-				first_notcon = 0;
-			} else
-				apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mp_apicid, pin);
-			continue;
-		}
-		if (!first_notcon) {
-			apic_printk(APIC_VERBOSE, " not connected.\n");
-			first_notcon = 1;
-		}
-
-		irq = pin_2_irq(idx, apic, pin);
-
-                if (multi_timer_check(apic, irq))
-                        continue;
-
-		add_pin_to_irq(irq, apic, pin);
-
-		setup_IO_APIC_irq(apic, pin, irq,
-				  irq_trigger(idx), irq_polarity(idx));
-	}
-	}
-
-	if (!first_notcon)
-		apic_printk(APIC_VERBOSE, " not connected.\n");
-}
-
-/*
- * Set up the timer pin, possibly with the 8259A-master behind.
- */
-static void __init setup_timer_IRQ0_pin(unsigned int apic, unsigned int pin,
-					int vector)
-{
-	struct IO_APIC_route_entry entry;
-
-	memset(&entry, 0, sizeof(entry));
-
-	/*
-	 * We use logical delivery to get the timer IRQ
-	 * to the first CPU.
-	 */
-	entry.dest_mode = INT_DEST_MODE;
-	entry.mask = 1;					/* mask IRQ now */
-	entry.dest = cpu_mask_to_apicid(TARGET_CPUS);
-	entry.delivery_mode = INT_DELIVERY_MODE;
-	entry.polarity = 0;
-	entry.trigger = 0;
-	entry.vector = vector;
-
-	/*
-	 * The timer IRQ doesn't have to know that behind the
-	 * scene we may have a 8259A-master in AEOI mode ...
-	 */
-	ioapic_register_intr(0, IOAPIC_EDGE);
-
-	/*
-	 * Add it to the IO-APIC irq-routing table:
-	 */
-	ioapic_write_entry(apic, pin, entry);
-}
-
-
-__apicdebuginit(void) print_IO_APIC(void)
-{
-	int apic, i;
-	union IO_APIC_reg_00 reg_00;
-	union IO_APIC_reg_01 reg_01;
-	union IO_APIC_reg_02 reg_02;
-	union IO_APIC_reg_03 reg_03;
-	unsigned long flags;
-	struct irq_cfg *cfg;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries);
-	for (i = 0; i < nr_ioapics; i++)
-		printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n",
-		       mp_ioapics[i].mp_apicid, nr_ioapic_registers[i]);
-
-	/*
-	 * We are a bit conservative about what we expect.  We have to
-	 * know about every hardware change ASAP.
-	 */
-	printk(KERN_INFO "testing the IO APIC.......................\n");
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	reg_00.raw = io_apic_read(apic, 0);
-	reg_01.raw = io_apic_read(apic, 1);
-	if (reg_01.bits.version >= 0x10)
-		reg_02.raw = io_apic_read(apic, 2);
-	if (reg_01.bits.version >= 0x20)
-		reg_03.raw = io_apic_read(apic, 3);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid);
-	printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
-	printk(KERN_DEBUG ".......    : physical APIC id: %02X\n", reg_00.bits.ID);
-	printk(KERN_DEBUG ".......    : Delivery Type: %X\n", reg_00.bits.delivery_type);
-	printk(KERN_DEBUG ".......    : LTS          : %X\n", reg_00.bits.LTS);
-
-	printk(KERN_DEBUG ".... register #01: %08X\n", reg_01.raw);
-	printk(KERN_DEBUG ".......     : max redirection entries: %04X\n", reg_01.bits.entries);
-
-	printk(KERN_DEBUG ".......     : PRQ implemented: %X\n", reg_01.bits.PRQ);
-	printk(KERN_DEBUG ".......     : IO APIC version: %04X\n", reg_01.bits.version);
-
-	/*
-	 * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02,
-	 * but the value of reg_02 is read as the previous read register
-	 * value, so ignore it if reg_02 == reg_01.
-	 */
-	if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) {
-		printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
-		printk(KERN_DEBUG ".......     : arbitration: %02X\n", reg_02.bits.arbitration);
-	}
-
-	/*
-	 * Some Intel chipsets with IO APIC VERSION of 0x2? don't have reg_02
-	 * or reg_03, but the value of reg_0[23] is read as the previous read
-	 * register value, so ignore it if reg_03 == reg_0[12].
-	 */
-	if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw &&
-	    reg_03.raw != reg_01.raw) {
-		printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw);
-		printk(KERN_DEBUG ".......     : Boot DT    : %X\n", reg_03.bits.boot_DT);
-	}
-
-	printk(KERN_DEBUG ".... IRQ redirection table:\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;
-
-		entry = ioapic_read_entry(apic, i);
-
-		printk(KERN_DEBUG " %02x %02X  ", i, entry.dest);
-
-		printk("%1d    %1d    %1d   %1d   %1d    %1d    %1d    %02X\n",
-			entry.mask,
-			entry.trigger,
-			entry.irr,
-			entry.polarity,
-			entry.delivery_status,
-			entry.dest_mode,
-			entry.delivery_mode,
-			entry.vector
-		);
-	}
-	}
-	printk(KERN_DEBUG "IRQ to pin mappings:\n");
-	for_each_irq_cfg(cfg) {
-		struct irq_pin_list *entry = cfg->irq_2_pin;
-		if (!entry)
-			continue;
-		printk(KERN_DEBUG "IRQ%d ", i);
-		for (;;) {
-			printk("-> %d:%d", entry->apic, entry->pin);
-			if (!entry->next)
-				break;
-			entry = entry->next;
-		}
-		printk("\n");
-	}
-
-	printk(KERN_INFO ".................................... done.\n");
-
-	return;
-}
-
-__apicdebuginit(void) print_APIC_bitfield(int base)
-{
-	unsigned int v;
-	int i, j;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
-	for (i = 0; i < 8; i++) {
-		v = apic_read(base + i*0x10);
-		for (j = 0; j < 32; j++) {
-			if (v & (1<<j))
-				printk("1");
-			else
-				printk("0");
-		}
-		printk("\n");
-	}
-}
-
-__apicdebuginit(void) print_local_APIC(void *dummy)
-{
-	unsigned int v, ver, maxlvt;
-	u64 icr;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
-		smp_processor_id(), hard_smp_processor_id());
-	v = apic_read(APIC_ID);
-	printk(KERN_INFO "... APIC ID:      %08x (%01x)\n", v,
-			GET_APIC_ID(v));
-	v = apic_read(APIC_LVR);
-	printk(KERN_INFO "... APIC VERSION: %08x\n", v);
-	ver = GET_APIC_VERSION(v);
-	maxlvt = lapic_get_maxlvt();
-
-	v = apic_read(APIC_TASKPRI);
-	printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
-
-	if (APIC_INTEGRATED(ver)) {			/* !82489DX */
-		v = apic_read(APIC_ARBPRI);
-		printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
-			v & APIC_ARBPRI_MASK);
-		v = apic_read(APIC_PROCPRI);
-		printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
-	}
-
-	v = apic_read(APIC_EOI);
-	printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
-	v = apic_read(APIC_RRR);
-	printk(KERN_DEBUG "... APIC RRR: %08x\n", v);
-	v = apic_read(APIC_LDR);
-	printk(KERN_DEBUG "... APIC LDR: %08x\n", v);
-	v = apic_read(APIC_DFR);
-	printk(KERN_DEBUG "... APIC DFR: %08x\n", v);
-	v = apic_read(APIC_SPIV);
-	printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
-
-	printk(KERN_DEBUG "... APIC ISR field:\n");
-	print_APIC_bitfield(APIC_ISR);
-	printk(KERN_DEBUG "... APIC TMR field:\n");
-	print_APIC_bitfield(APIC_TMR);
-	printk(KERN_DEBUG "... APIC IRR field:\n");
-	print_APIC_bitfield(APIC_IRR);
-
-	if (APIC_INTEGRATED(ver)) {		/* !82489DX */
-		if (maxlvt > 3)		/* Due to the Pentium erratum 3AP. */
-			apic_write(APIC_ESR, 0);
-		v = apic_read(APIC_ESR);
-		printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
-	}
-
-	icr = apic_icr_read();
-	printk(KERN_DEBUG "... APIC ICR: %08x\n", icr);
-	printk(KERN_DEBUG "... APIC ICR2: %08x\n", icr >> 32);
-
-	v = apic_read(APIC_LVTT);
-	printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
-
-	if (maxlvt > 3) {                       /* PC is LVT#4. */
-		v = apic_read(APIC_LVTPC);
-		printk(KERN_DEBUG "... APIC LVTPC: %08x\n", v);
-	}
-	v = apic_read(APIC_LVT0);
-	printk(KERN_DEBUG "... APIC LVT0: %08x\n", v);
-	v = apic_read(APIC_LVT1);
-	printk(KERN_DEBUG "... APIC LVT1: %08x\n", v);
-
-	if (maxlvt > 2) {			/* ERR is LVT#3. */
-		v = apic_read(APIC_LVTERR);
-		printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
-	}
-
-	v = apic_read(APIC_TMICT);
-	printk(KERN_DEBUG "... APIC TMICT: %08x\n", v);
-	v = apic_read(APIC_TMCCT);
-	printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
-	v = apic_read(APIC_TDCR);
-	printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
-	printk("\n");
-}
-
-__apicdebuginit(void) print_all_local_APICs(void)
-{
-	on_each_cpu(print_local_APIC, NULL, 1);
-}
-
-__apicdebuginit(void) print_PIC(void)
-{
-	unsigned int v;
-	unsigned long flags;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk(KERN_DEBUG "\nprinting PIC contents\n");
-
-	spin_lock_irqsave(&i8259A_lock, flags);
-
-	v = inb(0xa1) << 8 | inb(0x21);
-	printk(KERN_DEBUG "... PIC  IMR: %04x\n", v);
-
-	v = inb(0xa0) << 8 | inb(0x20);
-	printk(KERN_DEBUG "... PIC  IRR: %04x\n", v);
-
-	outb(0x0b, 0xa0);
-	outb(0x0b, 0x20);
-	v = inb(0xa0) << 8 | inb(0x20);
-	outb(0x0a, 0xa0);
-	outb(0x0a, 0x20);
-
-	spin_unlock_irqrestore(&i8259A_lock, flags);
-
-	printk(KERN_DEBUG "... PIC  ISR: %04x\n", v);
-
-	v = inb(0x4d1) << 8 | inb(0x4d0);
-	printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
-}
-
-__apicdebuginit(int) print_all_ICs(void)
-{
-	print_PIC();
-	print_all_local_APICs();
-	print_IO_APIC();
-
-	return 0;
-}
-
-fs_initcall(print_all_ICs);
-
-
-/* Where if anywhere is the i8259 connect in external int mode */
-static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
-
-static void __init enable_IO_APIC(void)
-{
-	union IO_APIC_reg_01 reg_01;
-	int i8259_apic, i8259_pin;
-	int i, apic;
-	unsigned long flags;
-
-	if (!pirqs_enabled)
-		for (i = 0; i < MAX_PIRQS; i++)
-			pirq_entries[i] = -1;
-
-	/*
-	 * The number of IO-APIC IRQ registers (== #pins):
-	 */
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		spin_lock_irqsave(&ioapic_lock, flags);
-		reg_01.raw = io_apic_read(apic, 1);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
-	}
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		int pin;
-		/* See if any of the pins is in ExtINT mode */
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-			struct IO_APIC_route_entry entry;
-			entry = ioapic_read_entry(apic, pin);
-
-			/* If the interrupt line is enabled and in ExtInt mode
-			 * I have found the pin where the i8259 is connected.
-			 */
-			if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) {
-				ioapic_i8259.apic = apic;
-				ioapic_i8259.pin  = pin;
-				goto found_i8259;
-			}
-		}
-	}
- found_i8259:
-	/* Look to see what if the MP table has reported the ExtINT */
-	/* If we could not find the appropriate pin by looking at the ioapic
-	 * the i8259 probably is not connected the ioapic but give the
-	 * mptable a chance anyway.
-	 */
-	i8259_pin  = find_isa_irq_pin(0, mp_ExtINT);
-	i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
-	/* Trust the MP table if nothing is setup in the hardware */
-	if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) {
-		printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n");
-		ioapic_i8259.pin  = i8259_pin;
-		ioapic_i8259.apic = i8259_apic;
-	}
-	/* Complain if the MP table and the hardware disagree */
-	if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) &&
-		(i8259_pin >= 0) && (ioapic_i8259.pin >= 0))
-	{
-		printk(KERN_WARNING "ExtINT in hardware and MP table differ\n");
-	}
-
-	/*
-	 * Do not trust the IO-APIC being empty at bootup
-	 */
-	clear_IO_APIC();
-}
-
-/*
- * Not an __init, needed by the reboot code
- */
-void disable_IO_APIC(void)
-{
-	/*
-	 * Clear the IO-APIC before rebooting:
-	 */
-	clear_IO_APIC();
-
-	/*
-	 * 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;
-
-		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	      = read_apic_id();
-
-		/*
-		 * Add it to the IO-APIC irq-routing table:
-		 */
-		ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry);
-	}
-	disconnect_bsp_APIC(ioapic_i8259.pin != -1);
-}
-
-/*
- * function to set the IO-APIC physical IDs based on the
- * values stored in the MPC table.
- *
- * by Matt Domsch <Matt_Domsch@dell.com>  Tue Dec 21 12:25:05 CST 1999
- */
-
-static void __init setup_ioapic_ids_from_mpc(void)
-{
-	union IO_APIC_reg_00 reg_00;
-	physid_mask_t phys_id_present_map;
-	int apic;
-	int i;
-	unsigned char old_id;
-	unsigned long flags;
-
-	if (x86_quirks->setup_ioapic_ids && x86_quirks->setup_ioapic_ids())
-		return;
-
-	/*
-	 * Don't check I/O APIC IDs for xAPIC systems.  They have
-	 * no meaning without the serial APIC bus.
-	 */
-	if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
-		|| APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
-		return;
-	/*
-	 * This is broken; anything with a real cpu count has to
-	 * circumvent this idiocy regardless.
-	 */
-	phys_id_present_map = ioapic_phys_id_map(phys_cpu_present_map);
-
-	/*
-	 * 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].mp_apicid;
-
-		if (mp_ioapics[apic].mp_apicid >= get_physical_broadcast()) {
-			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n",
-				apic, mp_ioapics[apic].mp_apicid);
-			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
-				reg_00.bits.ID);
-			mp_ioapics[apic].mp_apicid = reg_00.bits.ID;
-		}
-
-		/*
-		 * Sanity check, is the ID really free? Every APIC in a
-		 * system must have a unique ID or we get lots of nice
-		 * 'stuck on smp_invalidate_needed IPI wait' messages.
-		 */
-		if (check_apicid_used(phys_id_present_map,
-					mp_ioapics[apic].mp_apicid)) {
-			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n",
-				apic, mp_ioapics[apic].mp_apicid);
-			for (i = 0; i < get_physical_broadcast(); i++)
-				if (!physid_isset(i, phys_id_present_map))
-					break;
-			if (i >= get_physical_broadcast())
-				panic("Max APIC ID exceeded!\n");
-			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
-				i);
-			physid_set(i, phys_id_present_map);
-			mp_ioapics[apic].mp_apicid = i;
-		} else {
-			physid_mask_t tmp;
-			tmp = apicid_to_cpu_present(mp_ioapics[apic].mp_apicid);
-			apic_printk(APIC_VERBOSE, "Setting %d in the "
-					"phys_id_present_map\n",
-					mp_ioapics[apic].mp_apicid);
-			physids_or(phys_id_present_map, phys_id_present_map, tmp);
-		}
-
-
-		/*
-		 * We need to adjust the IRQ routing table
-		 * if the ID changed.
-		 */
-		if (old_id != mp_ioapics[apic].mp_apicid)
-			for (i = 0; i < mp_irq_entries; i++)
-				if (mp_irqs[i].mp_dstapic == old_id)
-					mp_irqs[i].mp_dstapic
-						= mp_ioapics[apic].mp_apicid;
-
-		/*
-		 * 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].mp_apicid);
-
-		reg_00.bits.ID = mp_ioapics[apic].mp_apicid;
-		spin_lock_irqsave(&ioapic_lock, flags);
-		io_apic_write(apic, 0, reg_00.raw);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-
-		/*
-		 * Sanity check
-		 */
-		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].mp_apicid)
-			printk("could not set ID!\n");
-		else
-			apic_printk(APIC_VERBOSE, " ok.\n");
-	}
-}
-
-int no_timer_check __initdata;
-
-static int __init notimercheck(char *s)
-{
-	no_timer_check = 1;
-	return 1;
-}
-__setup("no_timer_check", notimercheck);
-
-/*
- * There is a nasty bug in some older SMP boards, their mptable lies
- * about the timer IRQ. We do the following to work around the situation:
- *
- *	- timer IRQ defaults to IO-APIC IRQ
- *	- if this function detects that timer IRQs are defunct, then we fall
- *	  back to ISA timer IRQs
- */
-static int __init timer_irq_works(void)
-{
-	unsigned long t1 = jiffies;
-	unsigned long flags;
-
-	if (no_timer_check)
-		return 1;
-
-	local_save_flags(flags);
-	local_irq_enable();
-	/* Let ten ticks pass... */
-	mdelay((10 * 1000) / HZ);
-	local_irq_restore(flags);
-
-	/*
-	 * Expect a few ticks at least, to be sure some possible
-	 * glue logic does not lock up after one or two first
-	 * ticks in a non-ExtINT mode.  Also the local APIC
-	 * might have cached one ExtINT interrupt.  Finally, at
-	 * least one tick may be lost due to delays.
-	 */
-	if (time_after(jiffies, t1 + 4))
-		return 1;
-
-	return 0;
-}
-
-/*
- * In the SMP+IOAPIC case it might happen that there are an unspecified
- * number of pending IRQ events unhandled. These cases are very rare,
- * so we 'resend' these IRQs via IPIs, to the same CPU. It's much
- * better to do it this way as thus we do not have to be aware of
- * 'pending' interrupts in the IRQ path, except at this point.
- */
-/*
- * Edge triggered needs to resend any interrupt
- * that was delayed but this is now handled in the device
- * independent code.
- */
-
-/*
- * Startup quirk:
- *
- * Starting up a edge-triggered IO-APIC interrupt is
- * nasty - we need to make sure that we get the edge.
- * If it is already asserted for some reason, we need
- * return 1 to indicate that is was pending.
- *
- * This is not complete - we should be able to fake
- * an edge even if it isn't on the 8259A...
- *
- * (We do this for level-triggered IRQs too - it cannot hurt.)
- */
-static unsigned int startup_ioapic_irq(unsigned int irq)
-{
-	int was_pending = 0;
-	unsigned long flags;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	if (irq < 16) {
-		disable_8259A_irq(irq);
-		if (i8259A_irq_pending(irq))
-			was_pending = 1;
-	}
-	__unmask_IO_APIC_irq(irq);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	return was_pending;
-}
-
-static int ioapic_retrigger_irq(unsigned int irq)
-{
-	send_IPI_self(irq_cfg(irq)->vector);
-
-	return 1;
-}
-
-#ifdef CONFIG_SMP
-asmlinkage void smp_irq_move_cleanup_interrupt(void)
-{
-	unsigned vector, me;
-	ack_APIC_irq();
-	irq_enter();
-
-	me = smp_processor_id();
-	for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
-		unsigned int irq;
-		struct irq_desc *desc;
-		struct irq_cfg *cfg;
-		irq = __get_cpu_var(vector_irq)[vector];
-
-		desc = irq_desc(irq);
-		if (!desc)
-			continue;
-
-		cfg = irq_cfg(irq);
-		spin_lock(&desc->lock);
-		if (!cfg->move_cleanup_count)
-			goto unlock;
-
-		if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
-			goto unlock;
-
-		__get_cpu_var(vector_irq)[vector] = -1;
-		cfg->move_cleanup_count--;
-unlock:
-		spin_unlock(&desc->lock);
-	}
-
-	irq_exit();
-}
-
-static void irq_complete_move(unsigned int irq)
-{
-	struct irq_cfg *cfg = irq_cfg(irq);
-	unsigned vector, me;
-
-	if (likely(!cfg->move_in_progress))
-		return;
-
-	vector = ~get_irq_regs()->orig_ax;
-	me = smp_processor_id();
-	if ((vector == cfg->vector) && cpu_isset(me, cfg->domain)) {
-		cpumask_t cleanup_mask;
-
-		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
-		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
-		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
-		cfg->move_in_progress = 0;
-	}
-}
-#else
-static inline void irq_complete_move(unsigned int irq) {}
-#endif
-
-static void ack_apic_edge(unsigned int irq, struct irq_desc *desc)
-{
-	irq_complete_move(irq);
-	move_native_irq(irq, desc);
-	ack_APIC_irq();
-}
-
-#ifdef CONFIG_X86_64
-static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
-{
-        int do_unmask_irq = 0;
-
-        irq_complete_move(irq);
-#ifdef CONFIG_GENERIC_PENDING_IRQ
-        /* If we are moving the irq we need to mask it */
-        if (unlikely(desc->status & IRQ_MOVE_PENDING)) {
-                do_unmask_irq = 1;
-                mask_IO_APIC_irq(irq);
-        }
-#endif
-
-        /*
-         * We must acknowledge the irq before we move it or the acknowledge will
-         * not propagate properly.
-         */
-        ack_APIC_irq();
-
-        /* Now we can move and renable the irq */
-        if (unlikely(do_unmask_irq)) {
-                /* Only migrate the irq if the ack has been received.
-                 *
-                 * On rare occasions the broadcast level triggered ack gets
-                 * delayed going to ioapics, and if we reprogram the
-                 * vector while Remote IRR is still set the irq will never
-                 * fire again.
-                 *
-                 * To prevent this scenario we read the Remote IRR bit
-                 * of the ioapic.  This has two effects.
-                 * - On any sane system the read of the ioapic will
-                 *   flush writes (and acks) going to the ioapic from
-                 *   this cpu.
-                 * - We get to see if the ACK has actually been delivered.
-                 *
-                 * Based on failed experiments of reprogramming the
-                 * ioapic entry from outside of irq context starting
-                 * with masking the ioapic entry and then polling until
-                 * Remote IRR was clear before reprogramming the
-                 * ioapic I don't trust the Remote IRR bit to be
-                 * completey accurate.
-                 *
-                 * However there appears to be no other way to plug
-                 * this race, so if the Remote IRR bit is not
-                 * accurate and is causing problems then it is a hardware bug
-                 * and you can go talk to the chipset vendor about it.
-                 */
-                if (!io_apic_level_ack_pending(irq))
-                        move_masked_irq(irq, desc);
-                unmask_IO_APIC_irq(irq);
-        }
-}
-#else
-atomic_t irq_mis_count;
-static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
-{
-	unsigned long v;
-	int i;
-
-	irq_complete_move(irq);
-	move_native_irq(irq, desc);
-/*
- * 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
- */
-	i = irq_cfg(irq)->vector;
-
-	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
-
-	ack_APIC_irq();
-
-	if (!(v & (1 << (i & 0x1f)))) {
-		atomic_inc(&irq_mis_count);
-		spin_lock(&ioapic_lock);
-		__mask_and_edge_IO_APIC_irq(irq);
-		__unmask_and_level_IO_APIC_irq(irq);
-		spin_unlock(&ioapic_lock);
-	}
-}
-#endif
-
-static struct irq_chip ioapic_chip __read_mostly = {
-	.name 		= "IO-APIC",
-	.startup 	= startup_ioapic_irq,
-	.mask	 	= mask_IO_APIC_irq,
-	.unmask	 	= unmask_IO_APIC_irq,
-	.ack 		= ack_apic_edge,
-	.eoi 		= ack_apic_level,
-#ifdef CONFIG_SMP
-	.set_affinity 	= set_ioapic_affinity_irq,
-#endif
-	.retrigger	= ioapic_retrigger_irq,
-};
-
-
-static inline void init_IO_APIC_traps(void)
-{
-	int irq;
-	struct irq_desc *desc;
-	struct irq_cfg *cfg;
-
-	/*
-	 * NOTE! The local APIC isn't very good at handling
-	 * multiple interrupts at the same interrupt level.
-	 * As the interrupt level is determined by taking the
-	 * vector number and shifting that right by 4, we
-	 * want to spread these out a bit so that they don't
-	 * all fall in the same interrupt level.
-	 *
-	 * Also, we've got to be careful not to trash gate
-	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
-	 */
-	for_each_irq_cfg(cfg) {
-		irq = cfg->irq;
-		if (IO_APIC_IRQ(irq) && !cfg->vector) {
-			/*
-			 * Hmm.. We don't have an entry for this,
-			 * so default to an old-fashioned 8259
-			 * interrupt if we can..
-			 */
-			if (irq < 16)
-				make_8259A_irq(irq);
-			else {
-				desc = irq_desc(irq);
-				/* Strange. Oh, well.. */
-				desc->chip = &no_irq_chip;
-			}
-		}
-	}
-}
-
-/*
- * The local APIC irq-chip implementation:
- */
-
-static void mask_lapic_irq(unsigned int irq)
-{
-	unsigned long v;
-
-	v = apic_read(APIC_LVT0);
-	apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
-}
-
-static void unmask_lapic_irq(unsigned int irq)
-{
-	unsigned long v;
-
-	v = apic_read(APIC_LVT0);
-	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
-}
-
-static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
-{
-	ack_APIC_irq();
-}
-
-static struct irq_chip lapic_chip __read_mostly = {
-	.name		= "local-APIC",
-	.mask		= mask_lapic_irq,
-	.unmask		= unmask_lapic_irq,
-	.ack		= ack_lapic_irq,
-};
-
-static void lapic_register_intr(int irq)
-{
-	struct irq_desc *desc;
-
-	desc = irq_desc(irq);
-	desc->status &= ~IRQ_LEVEL;
-	set_irq_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq,
-				      "edge");
-}
-
-static void __init setup_nmi(void)
-{
-	/*
-	 * Dirty trick to enable the NMI watchdog ...
-	 * We put the 8259A master into AEOI mode and
-	 * unmask on all local APICs LVT0 as NMI.
-	 *
-	 * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
-	 * is from Maciej W. Rozycki - so we do not have to EOI from
-	 * the NMI handler or the timer interrupt.
-	 */
-	apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ...");
-
-	enable_NMI_through_LVT0();
-
-	apic_printk(APIC_VERBOSE, " done.\n");
-}
-
-/*
- * This looks a bit hackish but it's about the only one way of sending
- * a few INTA cycles to 8259As and any associated glue logic.  ICR does
- * not support the ExtINT mode, unfortunately.  We need to send these
- * cycles as some i82489DX-based boards have glue logic that keeps the
- * 8259A interrupt line asserted until INTA.  --macro
- */
-static inline void __init unlock_ExtINT_logic(void)
-{
-	int apic, pin, i;
-	struct IO_APIC_route_entry entry0, entry1;
-	unsigned char save_control, save_freq_select;
-
-	pin  = find_isa_irq_pin(8, mp_INT);
-	if (pin == -1) {
-		WARN_ON_ONCE(1);
-		return;
-	}
-	apic = find_isa_irq_apic(8, mp_INT);
-	if (apic == -1) {
-		WARN_ON_ONCE(1);
-		return;
-	}
-
-	entry0 = ioapic_read_entry(apic, pin);
-	clear_IO_APIC_pin(apic, pin);
-
-	memset(&entry1, 0, sizeof(entry1));
-
-	entry1.dest_mode = 0;			/* physical delivery */
-	entry1.mask = 0;			/* unmask IRQ now */
-	entry1.dest = hard_smp_processor_id();
-	entry1.delivery_mode = dest_ExtINT;
-	entry1.polarity = entry0.polarity;
-	entry1.trigger = 0;
-	entry1.vector = 0;
-
-	ioapic_write_entry(apic, pin, entry1);
-
-	save_control = CMOS_READ(RTC_CONTROL);
-	save_freq_select = CMOS_READ(RTC_FREQ_SELECT);
-	CMOS_WRITE((save_freq_select & ~RTC_RATE_SELECT) | 0x6,
-		   RTC_FREQ_SELECT);
-	CMOS_WRITE(save_control | RTC_PIE, RTC_CONTROL);
-
-	i = 100;
-	while (i-- > 0) {
-		mdelay(10);
-		if ((CMOS_READ(RTC_INTR_FLAGS) & RTC_PF) == RTC_PF)
-			i -= 10;
-	}
-
-	CMOS_WRITE(save_control, RTC_CONTROL);
-	CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
-	clear_IO_APIC_pin(apic, pin);
-
-	ioapic_write_entry(apic, pin, entry0);
-}
-
-static int disable_timer_pin_1 __initdata;
-/* Actually the next is obsolete, but keep it for paranoid reasons -AK */
-static int __init parse_disable_timer_pin_1(char *arg)
-{
-	disable_timer_pin_1 = 1;
-	return 0;
-}
-early_param("disable_timer_pin_1", parse_disable_timer_pin_1);
-
-int timer_through_8259 __initdata;
-
-/*
- * This code may look a bit paranoid, but it's supposed to cooperate with
- * a wide range of boards and BIOS bugs.  Fortunately only the timer IRQ
- * is so screwy.  Thanks to Brian Perkins for testing/hacking this beast
- * fanatically on his truly buggy board.
- */
-static inline void __init check_timer(void)
-{
-	struct irq_cfg *cfg = irq_cfg(0);
-	int apic1, pin1, apic2, pin2;
-	unsigned long flags;
-	unsigned int ver;
-	int no_pin1 = 0;
-
-	local_irq_save(flags);
-
-	ver = apic_read(APIC_LVR);
-	ver = GET_APIC_VERSION(ver);
-
-	/*
-	 * get/set the timer IRQ vector:
-	 */
-	disable_8259A_irq(0);
-	assign_irq_vector(0, TARGET_CPUS);
-
-	/*
-	 * As IRQ0 is to be enabled in the 8259A, the virtual
-	 * wire has to be disabled in the local APIC.  Also
-	 * timer interrupts need to be acknowledged manually in
-	 * the 8259A for the i82489DX when using the NMI
-	 * watchdog as that APIC treats NMIs as level-triggered.
-	 * The AEOI mode will finish them in the 8259A
-	 * automatically.
-	 */
-	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
-	init_8259A(1);
-	timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver));
-
-	pin1  = find_isa_irq_pin(0, mp_INT);
-	apic1 = find_isa_irq_apic(0, mp_INT);
-	pin2  = ioapic_i8259.pin;
-	apic2 = ioapic_i8259.apic;
-
-	apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X "
-		    "apic1=%d pin1=%d apic2=%d pin2=%d\n",
-		    cfg->vector, apic1, pin1, apic2, pin2);
-
-	/*
-	 * Some BIOS writers are clueless and report the ExtINTA
-	 * I/O APIC input from the cascaded 8259A as the timer
-	 * interrupt input.  So just in case, if only one pin
-	 * was found above, try it both directly and through the
-	 * 8259A.
-	 */
-	if (pin1 == -1) {
-		pin1 = pin2;
-		apic1 = apic2;
-		no_pin1 = 1;
-	} else if (pin2 == -1) {
-		pin2 = pin1;
-		apic2 = apic1;
-	}
-
-	if (pin1 != -1) {
-		/*
-		 * Ok, does IRQ0 through the IOAPIC work?
-		 */
-		if (no_pin1) {
-			add_pin_to_irq(0, apic1, pin1);
-			setup_timer_IRQ0_pin(apic1, pin1, cfg->vector);
-		}
-		unmask_IO_APIC_irq(0);
-		if (timer_irq_works()) {
-			if (nmi_watchdog == NMI_IO_APIC) {
-				setup_nmi();
-				enable_8259A_irq(0);
-			}
-			if (disable_timer_pin_1 > 0)
-				clear_IO_APIC_pin(0, pin1);
-			goto out;
-		}
-		clear_IO_APIC_pin(apic1, pin1);
-		if (!no_pin1)
-			apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
-				    "8254 timer not connected to IO-APIC\n");
-
-		apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer "
-			    "(IRQ0) through the 8259A ...\n");
-		apic_printk(APIC_QUIET, KERN_INFO
-			    "..... (found apic %d pin %d) ...\n", apic2, pin2);
-		/*
-		 * legacy devices should be connected to IO APIC #0
-		 */
-		replace_pin_at_irq(0, apic1, pin1, apic2, pin2);
-		setup_timer_IRQ0_pin(apic2, pin2, cfg->vector);
-		unmask_IO_APIC_irq(0);
-		enable_8259A_irq(0);
-		if (timer_irq_works()) {
-			apic_printk(APIC_QUIET, KERN_INFO "....... works.\n");
-			timer_through_8259 = 1;
-			if (nmi_watchdog == NMI_IO_APIC) {
-				disable_8259A_irq(0);
-				setup_nmi();
-				enable_8259A_irq(0);
-			}
-			goto out;
-		}
-		/*
-		 * Cleanup, just in case ...
-		 */
-		disable_8259A_irq(0);
-		clear_IO_APIC_pin(apic2, pin2);
-		apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n");
-	}
-
-	if (nmi_watchdog == NMI_IO_APIC) {
-		apic_printk(APIC_QUIET, KERN_WARNING "timer doesn't work "
-			    "through the IO-APIC - disabling NMI Watchdog!\n");
-		nmi_watchdog = NMI_NONE;
-	}
-	timer_ack = 0;
-
-	apic_printk(APIC_QUIET, KERN_INFO
-		    "...trying to set up timer as Virtual Wire IRQ...\n");
-
-	lapic_register_intr(0);
-	apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector);	/* Fixed mode */
-	enable_8259A_irq(0);
-
-	if (timer_irq_works()) {
-		apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
-		goto out;
-	}
-	disable_8259A_irq(0);
-	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector);
-	apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n");
-
-	apic_printk(APIC_QUIET, KERN_INFO
-		    "...trying to set up timer as ExtINT IRQ...\n");
-
-	init_8259A(0);
-	make_8259A_irq(0);
-	apic_write(APIC_LVT0, APIC_DM_EXTINT);
-
-	unlock_ExtINT_logic();
-
-	if (timer_irq_works()) {
-		apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
-		goto out;
-	}
-	apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n");
-	panic("IO-APIC + timer doesn't work!  Boot with apic=debug and send a "
-		"report.  Then try booting with the 'noapic' option.\n");
-out:
-	local_irq_restore(flags);
-}
-
-/*
- * Traditionally ISA IRQ2 is the cascade IRQ, and is not available
- * to devices.  However there may be an I/O APIC pin available for
- * this interrupt regardless.  The pin may be left unconnected, but
- * typically it will be reused as an ExtINT cascade interrupt for
- * the master 8259A.  In the MPS case such a pin will normally be
- * reported as an ExtINT interrupt in the MP table.  With ACPI
- * there is no provision for ExtINT interrupts, and in the absence
- * of an override it would be treated as an ordinary ISA I/O APIC
- * interrupt, that is edge-triggered and unmasked by default.  We
- * used to do this, but it caused problems on some systems because
- * of the NMI watchdog and sometimes IRQ0 of the 8254 timer using
- * the same ExtINT cascade interrupt to drive the local APIC of the
- * bootstrap processor.  Therefore we refrain from routing IRQ2 to
- * the I/O APIC in all cases now.  No actual device should request
- * it anyway.  --macro
- */
-#define PIC_IRQS	(1 << PIC_CASCADE_IR)
-
-void __init setup_IO_APIC(void)
-{
-	enable_IO_APIC();
-
-	io_apic_irqs = ~PIC_IRQS;
-
-	printk("ENABLING IO-APIC IRQs\n");
-
-	/*
-	 * Set up IO-APIC IRQ routing.
-	 */
-	if (!acpi_ioapic)
-		setup_ioapic_ids_from_mpc();
-	sync_Arb_IDs();
-	setup_IO_APIC_irqs();
-	init_IO_APIC_traps();
-	check_timer();
-}
-
-/*
- *	Called after all the initialization is done. If we didnt find any
- *	APIC bugs then we can allow the modify fast path
- */
-
-static int __init io_apic_bug_finalize(void)
-{
-	if (sis_apic_bug == -1)
-		sis_apic_bug = 0;
-	return 0;
-}
-
-late_initcall(io_apic_bug_finalize);
-
-struct sysfs_ioapic_data {
-	struct sys_device dev;
-	struct IO_APIC_route_entry entry[0];
-};
-static struct sysfs_ioapic_data *mp_ioapic_data[MAX_IO_APICS];
-
-static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
-{
-	struct IO_APIC_route_entry *entry;
-	struct sysfs_ioapic_data *data;
-	int i;
-
-	data = container_of(dev, struct sysfs_ioapic_data, dev);
-	entry = data->entry;
-	for (i = 0; i < nr_ioapic_registers[dev->id]; i++)
-		entry[i] = ioapic_read_entry(dev->id, i);
-
-	return 0;
-}
-
-static int ioapic_resume(struct sys_device *dev)
-{
-	struct IO_APIC_route_entry *entry;
-	struct sysfs_ioapic_data *data;
-	unsigned long flags;
-	union IO_APIC_reg_00 reg_00;
-	int i;
-
-	data = container_of(dev, struct sysfs_ioapic_data, dev);
-	entry = data->entry;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	reg_00.raw = io_apic_read(dev->id, 0);
-	if (reg_00.bits.ID != mp_ioapics[dev->id].mp_apicid) {
-		reg_00.bits.ID = mp_ioapics[dev->id].mp_apicid;
-		io_apic_write(dev->id, 0, reg_00.raw);
-	}
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-	for (i = 0; i < nr_ioapic_registers[dev->id]; i++)
-		ioapic_write_entry(dev->id, i, entry[i]);
-
-	return 0;
-}
-
-static struct sysdev_class ioapic_sysdev_class = {
-	.name = "ioapic",
-	.suspend = ioapic_suspend,
-	.resume = ioapic_resume,
-};
-
-static int __init ioapic_init_sysfs(void)
-{
-	struct sys_device *dev;
-	int i, size, error = 0;
-
-	error = sysdev_class_register(&ioapic_sysdev_class);
-	if (error)
-		return error;
-
-	for (i = 0; i < nr_ioapics; i++) {
-		size = sizeof(struct sys_device) + nr_ioapic_registers[i]
-			* sizeof(struct IO_APIC_route_entry);
-		mp_ioapic_data[i] = kzalloc(size, GFP_KERNEL);
-		if (!mp_ioapic_data[i]) {
-			printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
-			continue;
-		}
-		dev = &mp_ioapic_data[i]->dev;
-		dev->id = i;
-		dev->cls = &ioapic_sysdev_class;
-		error = sysdev_register(dev);
-		if (error) {
-			kfree(mp_ioapic_data[i]);
-			mp_ioapic_data[i] = NULL;
-			printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
-			continue;
-		}
-	}
-
-	return 0;
-}
-
-device_initcall(ioapic_init_sysfs);
-
-/*
- * Dynamic irq allocate and deallocation
- */
-unsigned int create_irq(unsigned int irq_want)
-{
-	/* Allocate an unused irq */
-	unsigned int irq, new;
-	unsigned long flags;
-	struct irq_cfg *cfg_new;
-
-#ifndef CONFIG_HAVE_SPARSE_IRQ
-	/* only can use bus/dev/fn.. when per_cpu vector is used */
-	irq_want = nr_irqs - 1;
-#endif
-
-	irq = 0;
-	spin_lock_irqsave(&vector_lock, flags);
-	for (new = irq_want; new > 0; new--) {
-		if (platform_legacy_irq(new))
-			continue;
-		cfg_new = irq_cfg(new);
-		if (cfg_new && cfg_new->vector != 0)
-			continue;
-		/* check if need to create one */
-		if (!cfg_new)
-			cfg_new = irq_cfg_with_new(new);
-		if (__assign_irq_vector(new, TARGET_CPUS) == 0)
-			irq = new;
-		break;
-	}
-	spin_unlock_irqrestore(&vector_lock, flags);
-
-	if (irq > 0) {
-		dynamic_irq_init(irq);
-	}
-	return irq;
-}
-
-void destroy_irq(unsigned int irq)
-{
-	unsigned long flags;
-
-	dynamic_irq_cleanup(irq);
-
-	spin_lock_irqsave(&vector_lock, flags);
-	__clear_irq_vector(irq);
-	spin_unlock_irqrestore(&vector_lock, flags);
-}
-
-/*
- * MSI message composition
- */
-#ifdef CONFIG_PCI_MSI
-static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
-{
-	struct irq_cfg *cfg;
-	int err;
-	unsigned dest;
-	cpumask_t tmp;
-
-	tmp = TARGET_CPUS;
-	err = assign_irq_vector(irq, tmp);
-	if (err)
-		return err;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, tmp);
-	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(cfg->vector);
-
-	return err;
-}
-
-#ifdef CONFIG_SMP
-static void set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	struct msi_msg msg;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	read_msi_msg(irq, &msg);
-
-	msg.data &= ~MSI_DATA_VECTOR_MASK;
-	msg.data |= MSI_DATA_VECTOR(cfg->vector);
-	msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
-	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
-
-	write_msi_msg(irq, &msg);
-	desc->affinity = 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,
-};
-
-
-static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc, int irq)
-{
-	int ret;
-	struct msi_msg msg;
-
-	ret = msi_compose_msg(dev, irq, &msg);
-	if (ret < 0)
-		return ret;
-
-	set_irq_msi(irq, desc);
-	write_msi_msg(irq, &msg);
-
-	set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
-
-	return 0;
-}
-
-static unsigned int build_irq_for_pci_dev(struct pci_dev *dev)
-{
-	unsigned int irq;
-
-	irq = dev->bus->number;
-	irq <<= 8;
-	irq |= dev->devfn;
-	irq <<= 12;
-
-	return irq;
-}
-
-int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
-{
-	int irq, ret;
-        unsigned int irq_want;
-
-        irq_want = build_irq_for_pci_dev(dev) + 0x100;
-
-        irq = create_irq(irq_want);
-
-	if (irq == 0)
-		return -1;
-
-	ret = setup_msi_irq(dev, desc, irq);
-	if (ret < 0) {
-		destroy_irq(irq);
-		return ret;
-        }
-
-	return 0;
-}
-
-int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
-{
-        unsigned int irq;
-        int ret, sub_handle;
-        struct msi_desc *desc;
-        unsigned int irq_want;
-
-        irq_want = build_irq_for_pci_dev(dev) + 0x100;
-        sub_handle = 0;
-        list_for_each_entry(desc, &dev->msi_list, list) {
-                irq = create_irq(irq_want--);
-                if (irq == 0)
-                        return -1;
-                ret = setup_msi_irq(dev, desc, irq);
-                if (ret < 0)
-                        goto error;
-                sub_handle++;
-        }
-        return 0;
-
-error:
-        destroy_irq(irq);
-        return ret;
-}
-
-
-void arch_teardown_msi_irq(unsigned int irq)
-{
-	destroy_irq(irq);
-}
-
-#endif /* CONFIG_PCI_MSI */
-
-/*
- * Hypertransport interrupt support
- */
-#ifdef CONFIG_HT_IRQ
-
-#ifdef CONFIG_SMP
-
-static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector)
-{
-	struct ht_irq_msg msg;
-	fetch_ht_irq_msg(irq, &msg);
-
-	msg.address_lo &= ~(HT_IRQ_LOW_VECTOR_MASK | HT_IRQ_LOW_DEST_ID_MASK);
-	msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK);
-
-	msg.address_lo |= HT_IRQ_LOW_VECTOR(vector) | HT_IRQ_LOW_DEST_ID(dest);
-	msg.address_hi |= HT_IRQ_HIGH_DEST_ID(dest);
-
-	write_ht_irq_msg(irq, &msg);
-}
-
-static void set_ht_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	target_ht_irq(irq, dest, cfg->vector);
-	desc->affinity = 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)
-{
-	struct irq_cfg *cfg;
-	int err;
-	cpumask_t tmp;
-
-	tmp = TARGET_CPUS;
-	err = assign_irq_vector(irq, tmp);
-	if ( !err) {
-		struct ht_irq_msg msg;
-		unsigned dest;
-
-		cfg = irq_cfg(irq);
-		cpus_and(tmp, cfg->domain, tmp);
-		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(cfg->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 err;
-}
-#endif /* CONFIG_HT_IRQ */
-
-/* --------------------------------------------------------------------------
-			ACPI-based IOAPIC Configuration
-   -------------------------------------------------------------------------- */
-
-#ifdef CONFIG_ACPI
-
-int __init io_apic_get_unique_id(int ioapic, int apic_id)
-{
-	union IO_APIC_reg_00 reg_00;
-	static physid_mask_t apic_id_map = PHYSID_MASK_NONE;
-	physid_mask_t tmp;
-	unsigned long flags;
-	int i = 0;
-
-	/*
-	 * The P4 platform supports up to 256 APIC IDs on two separate APIC
-	 * buses (one for LAPICs, one for IOAPICs), where predecessors only
-	 * supports up to 16 on one shared APIC bus.
-	 *
-	 * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full
-	 *      advantage of new APIC bus architecture.
-	 */
-
-	if (physids_empty(apic_id_map))
-		apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	reg_00.raw = io_apic_read(ioapic, 0);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	if (apic_id >= get_physical_broadcast()) {
-		printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying "
-			"%d\n", ioapic, apic_id, reg_00.bits.ID);
-		apic_id = reg_00.bits.ID;
-	}
-
-	/*
-	 * Every APIC in a system must have a unique ID or we get lots of nice
-	 * 'stuck on smp_invalidate_needed IPI wait' messages.
-	 */
-	if (check_apicid_used(apic_id_map, apic_id)) {
-
-		for (i = 0; i < get_physical_broadcast(); i++) {
-			if (!check_apicid_used(apic_id_map, i))
-				break;
-		}
-
-		if (i == get_physical_broadcast())
-			panic("Max apic_id exceeded!\n");
-
-		printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
-			"trying %d\n", ioapic, apic_id, i);
-
-		apic_id = i;
-	}
-
-	tmp = apicid_to_cpu_present(apic_id);
-	physids_or(apic_id_map, apic_id_map, tmp);
-
-	if (reg_00.bits.ID != apic_id) {
-		reg_00.bits.ID = apic_id;
-
-		spin_lock_irqsave(&ioapic_lock, flags);
-		io_apic_write(ioapic, 0, reg_00.raw);
-		reg_00.raw = io_apic_read(ioapic, 0);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-
-		/* Sanity check */
-		if (reg_00.bits.ID != apic_id) {
-			printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
-			return -1;
-		}
-	}
-
-	apic_printk(APIC_VERBOSE, KERN_INFO
-			"IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
-
-	return apic_id;
-}
-
-
-int __init io_apic_get_version(int ioapic)
-{
-	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);
-
-	return reg_01.bits.version;
-}
-
-
-int __init io_apic_get_redir_entries(int ioapic)
-{
-	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);
-
-	return reg_01.bits.entries;
-}
-
-
-int io_apic_set_pci_routing(int ioapic, int pin, int irq, int triggering, int polarity)
-{
-	if (!IO_APIC_IRQ(irq)) {
-		printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
-			ioapic);
-		return -EINVAL;
-	}
-
-	/*
-	 * IRQs < 16 are already in the irq_2_pin[] map
-	 */
-	if (irq >= 16)
-		add_pin_to_irq(irq, ioapic, pin);
-
-	setup_IO_APIC_irq(ioapic, pin, irq, triggering, polarity);
-
-	return 0;
-}
-
-int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
-{
-	int i;
-
-	if (skip_ioapic_setup)
-		return -1;
-
-	for (i = 0; i < mp_irq_entries; i++)
-		if (mp_irqs[i].mp_irqtype == mp_INT &&
-		    mp_irqs[i].mp_srcbusirq == bus_irq)
-			break;
-	if (i >= mp_irq_entries)
-		return -1;
-
-	*trigger = irq_trigger(i);
-	*polarity = irq_polarity(i);
-	return 0;
-}
-
-#endif /* CONFIG_ACPI */
-
-/*
- * This function currently is only a helper for the i386 smp boot process where
- * we need to reprogram the ioredtbls to cater for the cpus which have come online
- * so mask in all cases should simply be TARGET_CPUS
- */
-#ifdef CONFIG_SMP
-void __init setup_ioapic_dest(void)
-{
-	int pin, ioapic, irq, irq_entry;
-	struct irq_cfg *cfg;
-	struct irq_desc *desc;
-
-	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.
-			 */
-			cfg = irq_cfg(irq);
-			if (!cfg->vector)
-				setup_IO_APIC_irq(ioapic, pin, irq,
-						  irq_trigger(irq_entry),
-						  irq_polarity(irq_entry));
-			else {
-				desc = irq_desc(irq);
-				set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
-			}
-		}
-
-	}
-}
-#endif
-
-void __init ioapic_init_mappings(void)
-{
-	unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
-	int i;
-
-	for (i = 0; i < nr_ioapics; i++) {
-		if (smp_found_config) {
-			ioapic_phys = mp_ioapics[i].mp_apicaddr;
-			if (!ioapic_phys) {
-				printk(KERN_ERR
-				       "WARNING: bogus zero IO-APIC "
-				       "address found in MPTABLE, "
-				       "disabling IO/APIC support!\n");
-				smp_found_config = 0;
-				skip_ioapic_setup = 1;
-				goto fake_ioapic_page;
-			}
-		} else {
-fake_ioapic_page:
-			ioapic_phys = (unsigned long)
-				      alloc_bootmem_pages(PAGE_SIZE);
-			ioapic_phys = __pa(ioapic_phys);
-		}
-		set_fixmap_nocache(idx, ioapic_phys);
-		printk(KERN_DEBUG "mapped IOAPIC to %08lx (%08lx)\n",
-		       __fix_to_virt(idx), ioapic_phys);
-		idx++;
-	}
-}
-
Index: linux-2.6/arch/x86/kernel/io_apic_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic_64.c
+++ /dev/null
@@ -1,3896 +0,0 @@
-/*
- *	Intel IO-APIC support for multi-Pentium hosts.
- *
- *	Copyright (C) 1997, 1998, 1999, 2000 Ingo Molnar, Hajnalka Szabo
- *
- *	Many thanks to Stig Venaas for trying out countless experimental
- *	patches and reporting/debugging problems patiently!
- *
- *	(c) 1999, Multiple IO-APIC support, developed by
- *	Ken-ichi Yaku <yaku@css1.kbnes.nec.co.jp> and
- *      Hidemi Kishimoto <kisimoto@css1.kbnes.nec.co.jp>,
- *	further tested and cleaned up by Zach Brown <zab@redhat.com>
- *	and Ingo Molnar <mingo@redhat.com>
- *
- *	Fixes
- *	Maciej W. Rozycki	:	Bits for genuine 82489DX APICs;
- *					thanks to Eric Gilmore
- *					and Rolf G. Tews
- *					for testing these extensively
- *	Paul Diefenbaugh	:	Added full ACPI support
- */
-
-#include <linux/mm.h>
-#include <linux/interrupt.h>
-#include <linux/init.h>
-#include <linux/delay.h>
-#include <linux/sched.h>
-#include <linux/pci.h>
-#include <linux/mc146818rtc.h>
-#include <linux/compiler.h>
-#include <linux/acpi.h>
-#include <linux/module.h>
-#include <linux/sysdev.h>
-#include <linux/msi.h>
-#include <linux/htirq.h>
-#include <linux/freezer.h>
-#include <linux/kthread.h>
-#include <linux/jiffies.h>	/* time_after() */
-#ifdef CONFIG_ACPI
-#include <acpi/acpi_bus.h>
-#endif
-#include <linux/bootmem.h>
-#include <linux/dmar.h>
-
-#include <asm/idle.h>
-#include <asm/io.h>
-#include <asm/smp.h>
-#include <asm/desc.h>
-#include <asm/proto.h>
-#include <asm/acpi.h>
-#include <asm/dma.h>
-#include <asm/timer.h>
-#include <asm/i8259.h>
-#include <asm/nmi.h>
-#include <asm/msidef.h>
-#include <asm/hypertransport.h>
-#include <asm/setup.h>
-#include <asm/irq_remapping.h>
-
-#include <mach_ipi.h>
-#include <mach_apic.h>
-#include <mach_apicdef.h>
-
-#define __apicdebuginit(type) static type __init
-
-/*
- *      Is the SiS APIC rmw bug present ?
- *      -1 = don't know, 0 = no, 1 = yes
- */
-int sis_apic_bug = -1;
-
-static DEFINE_SPINLOCK(ioapic_lock);
-static DEFINE_SPINLOCK(vector_lock);
-
-int first_free_entry;
-/*
- * Rough estimation of how many shared IRQs there are, can
- * be changed anytime.
- */
-int pin_map_size;
-
-/*
- * # of IRQ routing registers
- */
-int nr_ioapic_registers[MAX_IO_APICS];
-
-/* I/O APIC entries */
-struct mp_config_ioapic mp_ioapics[MAX_IO_APICS];
-int nr_ioapics;
-
-/* MP IRQ source entries */
-struct mp_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
-
-/* # of MP IRQ source entries */
-int mp_irq_entries;
-
-#if defined (CONFIG_MCA) || defined (CONFIG_EISA)
-int mp_bus_id_to_type[MAX_MP_BUSSES];
-#endif
-
-DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES);
-
-int skip_ioapic_setup;
-
-static int __init parse_noapic(char *str)
-{
-	/* disable IO-APIC */
-	disable_ioapic_setup();
-	return 0;
-}
-early_param("noapic", parse_noapic);
-
-struct irq_cfg;
-struct irq_pin_list;
-struct irq_cfg {
-	unsigned int irq;
-	struct irq_cfg *next;
-	struct irq_pin_list *irq_2_pin;
-	cpumask_t domain;
-	cpumask_t old_domain;
-	unsigned move_cleanup_count;
-	u8 vector;
-	u8 move_in_progress : 1;
-};
-
-/* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
-static struct irq_cfg irq_cfg_legacy[] __initdata = {
-	[0]  = { .irq =  0, .domain = CPU_MASK_ALL, .vector = IRQ0_VECTOR,  },
-	[1]  = { .irq =  1, .domain = CPU_MASK_ALL, .vector = IRQ1_VECTOR,  },
-	[2]  = { .irq =  2, .domain = CPU_MASK_ALL, .vector = IRQ2_VECTOR,  },
-	[3]  = { .irq =  3, .domain = CPU_MASK_ALL, .vector = IRQ3_VECTOR,  },
-	[4]  = { .irq =  4, .domain = CPU_MASK_ALL, .vector = IRQ4_VECTOR,  },
-	[5]  = { .irq =  5, .domain = CPU_MASK_ALL, .vector = IRQ5_VECTOR,  },
-	[6]  = { .irq =  6, .domain = CPU_MASK_ALL, .vector = IRQ6_VECTOR,  },
-	[7]  = { .irq =  7, .domain = CPU_MASK_ALL, .vector = IRQ7_VECTOR,  },
-	[8]  = { .irq =  8, .domain = CPU_MASK_ALL, .vector = IRQ8_VECTOR,  },
-	[9]  = { .irq =  9, .domain = CPU_MASK_ALL, .vector = IRQ9_VECTOR,  },
-	[10] = { .irq = 10, .domain = CPU_MASK_ALL, .vector = IRQ10_VECTOR, },
-	[11] = { .irq = 11, .domain = CPU_MASK_ALL, .vector = IRQ11_VECTOR, },
-	[12] = { .irq = 12, .domain = CPU_MASK_ALL, .vector = IRQ12_VECTOR, },
-	[13] = { .irq = 13, .domain = CPU_MASK_ALL, .vector = IRQ13_VECTOR, },
-	[14] = { .irq = 14, .domain = CPU_MASK_ALL, .vector = IRQ14_VECTOR, },
-	[15] = { .irq = 15, .domain = CPU_MASK_ALL, .vector = IRQ15_VECTOR, },
-};
-
-static struct irq_cfg irq_cfg_init = { .irq =  -1U, };
-/* need to be biger than size of irq_cfg_legacy */
-static int nr_irq_cfg = 32;
-
-static int __init parse_nr_irq_cfg(char *arg)
-{
-	if (arg) {
-		nr_irq_cfg = simple_strtoul(arg, NULL, 0);
-		if (nr_irq_cfg < 32)
-			nr_irq_cfg = 32;
-	}
-	return 0;
-}
-
-early_param("nr_irq_cfg", parse_nr_irq_cfg);
-
-static void init_one_irq_cfg(struct irq_cfg *cfg)
-{
-	memcpy(cfg, &irq_cfg_init, sizeof(struct irq_cfg));
-}
-
-static struct irq_cfg *irq_cfgx;
-static struct irq_cfg *irq_cfgx_free;
-static void __init init_work(void *data)
-{
-	struct dyn_array *da = data;
-	struct irq_cfg *cfg;
-	int legacy_count;
-	int i;
-
-	cfg = *da->name;
-
-	memcpy(cfg, irq_cfg_legacy, sizeof(irq_cfg_legacy));
-
-	legacy_count = sizeof(irq_cfg_legacy)/sizeof(irq_cfg_legacy[0]);
-	for (i = legacy_count; i < *da->nr; i++)
-		init_one_irq_cfg(&cfg[i]);
-
-	for (i = 1; i < *da->nr; i++)
-		cfg[i-1].next = &cfg[i];
-
-	irq_cfgx_free = &irq_cfgx[legacy_count];
-	irq_cfgx[legacy_count - 1].next = NULL;
-}
-
-#define for_each_irq_cfg(cfg)		\
-	for (cfg = irq_cfgx; cfg; cfg = cfg->next)
-
-DEFINE_DYN_ARRAY(irq_cfgx, sizeof(struct irq_cfg), nr_irq_cfg, PAGE_SIZE, init_work);
-
-static struct irq_cfg *irq_cfg(unsigned int irq)
-{
-	struct irq_cfg *cfg;
-
-	cfg = irq_cfgx;
-	while (cfg) {
-		if (cfg->irq == irq)
-			return cfg;
-
-		cfg = cfg->next;
-	}
-
-	return NULL;
-}
-
-static struct irq_cfg *irq_cfg_with_new(unsigned int irq)
-{
-	struct irq_cfg *cfg, *cfg_pri;
-	int i;
-	int count = 0;
-
-	cfg_pri = cfg = irq_cfgx;
-	while (cfg) {
-		if (cfg->irq == irq)
-			return cfg;
-
-		cfg_pri = cfg;
-		cfg = cfg->next;
-		count++;
-	}
-
-	if (!irq_cfgx_free) {
-		unsigned long phys;
-		unsigned long total_bytes;
-		/*
-		 *  we run out of pre-allocate ones, allocate more
-		 */
-		printk(KERN_DEBUG "try to get more irq_cfg %d\n", nr_irq_cfg);
-
-		total_bytes = sizeof(struct irq_cfg) * nr_irq_cfg;
-		if (after_bootmem)
-			cfg = kzalloc(total_bytes, GFP_ATOMIC);
-		else
-			cfg = __alloc_bootmem_nopanic(total_bytes, PAGE_SIZE, 0);
-
-		if (!cfg)
-			panic("please boot with nr_irq_cfg= %d\n", count * 2);
-
-		phys = __pa(cfg);
-		printk(KERN_DEBUG "irq_irq ==> [%#lx - %#lx]\n", phys, phys + total_bytes);
-
-		for (i = 0; i < nr_irq_cfg; i++)
-			init_one_irq_cfg(&cfg[i]);
-
-		for (i = 1; i < nr_irq_cfg; i++)
-			cfg[i-1].next = &cfg[i];
-
-		irq_cfgx_free = cfg;
-	}
-
-	cfg = irq_cfgx_free;
-	irq_cfgx_free = irq_cfgx_free->next;
-	cfg->next = NULL;
-	if (cfg_pri)
-		cfg_pri->next = cfg;
-	else
-		irq_cfgx = cfg;
-	cfg->irq = irq;
-	printk(KERN_DEBUG "found new irq_cfg for irq %d\n", cfg->irq);
-#ifdef CONFIG_HAVE_SPARSE_IRQ_DEBUG
-	{
-		/* dump the results */
-		struct irq_cfg *cfg;
-		unsigned long phys;
-		unsigned long bytes = sizeof(struct irq_cfg);
-
-		printk(KERN_DEBUG "=========================== %d\n", irq);
-		printk(KERN_DEBUG "irq_cfg dump after get that for %d\n", irq);
-		for_each_irq_cfg(cfg) {
-			phys = __pa(cfg);
-			printk(KERN_DEBUG "irq_cfg %d ==> [%#lx - %#lx]\n", cfg->irq, phys, phys + bytes);
-		}
-		printk(KERN_DEBUG "===========================\n");
-	}
-#endif
-	return cfg;
-}
-
-/*
- * 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 *irq_2_pin_head;
-/* fill one page ? */
-static int nr_irq_2_pin = 0x100;
-static struct irq_pin_list *irq_2_pin_ptr;
-static void __init irq_2_pin_init_work(void *data)
-{
-	struct dyn_array *da = data;
-	struct irq_pin_list *pin;
-	int i;
-
-	pin = *da->name;
-
-	for (i = 1; i < *da->nr; i++)
-		pin[i-1].next = &pin[i];
-
-	irq_2_pin_ptr = &pin[0];
-}
-DEFINE_DYN_ARRAY(irq_2_pin_head, sizeof(struct irq_pin_list), nr_irq_2_pin, PAGE_SIZE, irq_2_pin_init_work);
-
-static struct irq_pin_list *get_one_free_irq_2_pin(void)
-{
-	struct irq_pin_list *pin;
-	int i;
-
-	pin = irq_2_pin_ptr;
-
-	if (pin) {
-		irq_2_pin_ptr = pin->next;
-		pin->next = NULL;
-		return pin;
-	}
-
-	/*
-	 *  we run out of pre-allocate ones, allocate more
-	 */
-	printk(KERN_DEBUG "try to get more irq_2_pin %d\n", nr_irq_2_pin);
-
-	if (after_bootmem)
-		pin = kzalloc(sizeof(struct irq_pin_list)*nr_irq_2_pin,
-				 GFP_ATOMIC);
-	else
-		pin = __alloc_bootmem_nopanic(sizeof(struct irq_pin_list) *
-				nr_irq_2_pin, PAGE_SIZE, 0);
-
-	if (!pin)
-		panic("can not get more irq_2_pin\n");
-
-	for (i = 1; i < nr_irq_2_pin; i++)
-		pin[i-1].next = &pin[i];
-
-	irq_2_pin_ptr = pin->next;
-	pin->next = NULL;
-
-	return pin;
-}
-
-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].mp_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.
- *
- * Older SiS APIC requires we rewrite the index register
- */
-static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
-{
-	struct io_apic __iomem *io_apic = io_apic_base(apic);
-        if (sis_apic_bug)
-                writel(reg, &io_apic->index);
-	writel(value, &io_apic->data);
-}
-
-#ifdef CONFIG_X86_64
-static bool io_apic_level_ack_pending(unsigned int irq)
-{
-	struct irq_pin_list *entry;
-	unsigned long flags;
-	struct irq_cfg *cfg = irq_cfg(irq);
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		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? */
-		if (reg & IO_APIC_REDIR_REMOTE_IRR) {
-			spin_unlock_irqrestore(&ioapic_lock, flags);
-			return true;
-		}
-		if (!entry->next)
-			break;
-		entry = entry->next;
-	}
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	return false;
-}
-#endif
-
-union entry_union {
-	struct { u32 w1, w2; };
-	struct IO_APIC_route_entry entry;
-};
-
-static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin)
-{
-	union entry_union eu;
-	unsigned long flags;
-	spin_lock_irqsave(&ioapic_lock, flags);
-	eu.w1 = io_apic_read(apic, 0x10 + 2 * pin);
-	eu.w2 = io_apic_read(apic, 0x11 + 2 * pin);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-	return eu.entry;
-}
-
-/*
- * When we write a new IO APIC routing entry, we need to write the high
- * word first! If the mask bit in the low word is clear, we will enable
- * the interrupt, and we need to make sure the entry is fully populated
- * before that happens.
- */
-static void
-__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
-{
-	union entry_union eu;
-	eu.entry = e;
-	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
-	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
-}
-
-static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
-{
-	unsigned long flags;
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__ioapic_write_entry(apic, pin, e);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-/*
- * When we mask an IO APIC routing entry, we need to write the low
- * word first, in order to set the mask bit before we change the
- * high bits!
- */
-static void ioapic_mask_entry(int apic, int pin)
-{
-	unsigned long flags;
-	union entry_union eu = { .entry.mask = 1 };
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
-	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-#ifdef CONFIG_SMP
-static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, u8 vector)
-{
-	int apic, pin;
-	struct irq_cfg *cfg;
-	struct irq_pin_list *entry;
-
-	cfg = irq_cfg(irq);
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		unsigned int reg;
-
-		if (!entry)
-			break;
-
-		apic = entry->apic;
-		pin = entry->pin;
-#ifdef CONFIG_INTR_REMAP
-		/*
-		 * With interrupt-remapping, destination information comes
-		 * from interrupt-remapping table entry.
-		 */
-		if (!irq_remapped(irq))
-			io_apic_write(apic, 0x11 + pin*2, dest);
-#else
-		io_apic_write(apic, 0x11 + pin*2, dest);
-#endif
-		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, cpumask_t mask);
-
-static void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	unsigned long flags;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	cfg = irq_cfg(irq);
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-	/*
-	 * Only the high 8 bits are valid.
-	 */
-	dest = SET_APIC_LOGICAL_ID(dest);
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__target_IO_APIC_irq(irq, dest, cfg->vector);
-	desc->affinity = mask;
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-#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(unsigned int irq, int apic, int pin)
-{
-	struct irq_cfg *cfg;
-	struct irq_pin_list *entry;
-
-	/* first time to refer irq_cfg, so with new */
-	cfg = irq_cfg_with_new(irq);
-	entry = cfg->irq_2_pin;
-	if (!entry) {
-		entry = get_one_free_irq_2_pin();
-		cfg->irq_2_pin = entry;
-		entry->apic = apic;
-		entry->pin = pin;
-		printk(KERN_DEBUG " 0 add_pin_to_irq: irq %d --> apic %d pin %d\n", irq, apic, pin);
-		return;
-	}
-
-	while (entry->next) {
-		/* not again, please */
-		if (entry->apic == apic && entry->pin == pin)
-			return;
-
-		entry = entry->next;
-	}
-
-	entry->next = get_one_free_irq_2_pin();
-	entry = entry->next;
-	entry->apic = apic;
-	entry->pin = pin;
-	printk(KERN_DEBUG " x add_pin_to_irq: irq %d --> apic %d pin %d\n", irq, apic, pin);
-}
-
-/*
- * Reroute an IRQ to a different pin.
- */
-static void __init replace_pin_at_irq(unsigned int irq,
-				      int oldapic, int oldpin,
-				      int newapic, int newpin)
-{
-	struct irq_cfg *cfg = irq_cfg(irq);
-	struct irq_pin_list *entry = cfg->irq_2_pin;
-	int replaced = 0;
-
-	while (entry) {
-		if (entry->apic == oldapic && entry->pin == oldpin) {
-			entry->apic = newapic;
-			entry->pin = newpin;
-			replaced = 1;
-			/* every one is different, right? */
-			break;
-		}
-		entry = entry->next;
-	}
-
-	/* why? call replace before add? */
-	if (!replaced)
-		add_pin_to_irq(irq, newapic, newpin);
-}
-
-#ifdef CONFIG_X86_64
-/*
- * Synchronize the IO-APIC and the CPU by doing
- * a dummy read from the IO-APIC
- */
-static inline void io_apic_sync(unsigned int apic)
-{
-	struct io_apic __iomem *io_apic = io_apic_base(apic);
-	readl(&io_apic->data);
-}
-
-#define __DO_ACTION(R, ACTION, FINAL)					\
-									\
-{									\
-	int pin;							\
-	struct irq_cfg *cfg;						\
-	struct irq_pin_list *entry;					\
-									\
-	cfg = irq_cfg(irq);						\
-	entry = cfg->irq_2_pin;						\
-	for (;;) {							\
-		unsigned int reg;					\
-		if (!entry)						\
-			break;						\
-		pin = entry->pin;					\
-		reg = io_apic_read(entry->apic, 0x10 + R + pin*2);	\
-		reg ACTION;						\
-		io_apic_modify(entry->apic, 0x10 + R + pin*2, reg);	\
-		FINAL;							\
-		if (!entry->next)					\
-			break;						\
-		entry = entry->next;					\
-	}								\
-}
-
-#define DO_ACTION(name,R,ACTION, FINAL)					\
-									\
-	static void name##_IO_APIC_irq (unsigned int irq)		\
-	__DO_ACTION(R, ACTION, FINAL)
-
-/* mask = 1 */
-DO_ACTION(__mask,	0, |= IO_APIC_REDIR_MASKED, io_apic_sync(entry->apic))
-
-/* mask = 0 */
-DO_ACTION(__unmask,	0, &= ~IO_APIC_REDIR_MASKED, )
-
-#else
-
-static void __modify_IO_APIC_irq(unsigned int irq, unsigned long enable, unsigned long disable)
-{
-	struct irq_cfg *cfg;
-	struct irq_pin_list *entry;
-	unsigned int pin, reg;
-
-	cfg = irq_cfg(irq);
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		if (!entry)
-			break;
-		pin = entry->pin;
-		reg = io_apic_read(entry->apic, 0x10 + pin*2);
-		reg &= ~disable;
-		reg |= enable;
-		io_apic_modify(entry->apic, 0x10 + pin*2, reg);
-		if (!entry->next)
-			break;
-		entry = entry->next;
-	}
-}
-
-/* mask = 1 */
-static void __mask_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED, 0);
-}
-
-/* mask = 0 */
-static void __unmask_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, 0, IO_APIC_REDIR_MASKED);
-}
-
-/* mask = 1, trigger = 0 */
-static void __mask_and_edge_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED,
-				IO_APIC_REDIR_LEVEL_TRIGGER);
-}
-
-/* mask = 0, trigger = 1 */
-static void __unmask_and_level_IO_APIC_irq(unsigned int irq)
-{
-	__modify_IO_APIC_irq(irq, IO_APIC_REDIR_LEVEL_TRIGGER,
-				IO_APIC_REDIR_MASKED);
-}
-
-#endif
-
-static void mask_IO_APIC_irq (unsigned int irq)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__mask_IO_APIC_irq(irq);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-static void unmask_IO_APIC_irq (unsigned int irq)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__unmask_IO_APIC_irq(irq);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
-{
-	struct IO_APIC_route_entry entry;
-
-	/* Check delivery_mode to be sure we're not clearing an SMI pin */
-	entry = ioapic_read_entry(apic, pin);
-	if (entry.delivery_mode == dest_SMI)
-		return;
-	/*
-	 * Disable it in the IO-APIC irq-routing table:
-	 */
-	ioapic_mask_entry(apic, pin);
-}
-
-static void clear_IO_APIC (void)
-{
-	int apic, pin;
-
-	for (apic = 0; apic < nr_ioapics; apic++)
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
-			clear_IO_APIC_pin(apic, pin);
-}
-
-#if !defined(CONFIG_SMP) && defined(CONFIG_X86_32)
-void send_IPI_self(int vector)
-{
-	unsigned int cfg;
-
-	/*
-	 * Wait for idle.
-	 */
-	apic_wait_icr_idle();
-	cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
-	/*
-	 * Send the IPI. The write to APIC_ICR fires this off.
-	 */
-	apic_write(APIC_ICR, cfg);
-}
-#endif /* !CONFIG_SMP && CONFIG_X86_32*/
-
-#ifdef CONFIG_X86_32
-/*
- * 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;
-
-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, KERN_INFO
-			"PIRQ redirection, working around broken MP-BIOS.\n");
-	max = MAX_PIRQS;
-	if (ints[0] < MAX_PIRQS)
-		max = ints[0];
-
-	for (i = 0; i < max; i++) {
-		apic_printk(APIC_VERBOSE, KERN_DEBUG
-				"... PIRQ%d -> IRQ %d\n", i, ints[i+1]);
-		/*
-		 * PIRQs are mapped upside down, usually.
-		 */
-		pirq_entries[MAX_PIRQS-i-1] = ints[i+1];
-	}
-	return 1;
-}
-
-__setup("pirq=", ioapic_pirq_setup);
-#endif /* CONFIG_X86_32 */
-
-#ifdef CONFIG_INTR_REMAP
-/* I/O APIC RTE contents at the OS boot up */
-static struct IO_APIC_route_entry *early_ioapic_entries[MAX_IO_APICS];
-
-/*
- * Saves and masks all the unmasked IO-APIC RTE's
- */
-int save_mask_IO_APIC_setup(void)
-{
-	union IO_APIC_reg_01 reg_01;
-	unsigned long flags;
-	int apic, pin;
-
-	/*
-	 * The number of IO-APIC IRQ registers (== #pins):
-	 */
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		spin_lock_irqsave(&ioapic_lock, flags);
-		reg_01.raw = io_apic_read(apic, 1);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
-	}
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		early_ioapic_entries[apic] =
-			kzalloc(sizeof(struct IO_APIC_route_entry) *
-				nr_ioapic_registers[apic], GFP_KERNEL);
-		if (!early_ioapic_entries[apic])
-			return -ENOMEM;
-	}
-
-	for (apic = 0; apic < nr_ioapics; apic++)
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-			struct IO_APIC_route_entry entry;
-
-			entry = early_ioapic_entries[apic][pin] =
-				ioapic_read_entry(apic, pin);
-			if (!entry.mask) {
-				entry.mask = 1;
-				ioapic_write_entry(apic, pin, entry);
-			}
-		}
-	return 0;
-}
-
-void restore_IO_APIC_setup(void)
-{
-	int apic, pin;
-
-	for (apic = 0; apic < nr_ioapics; apic++)
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
-			ioapic_write_entry(apic, pin,
-					   early_ioapic_entries[apic][pin]);
-}
-
-void reinit_intr_remapped_IO_APIC(int intr_remapping)
-{
-	/*
-	 * 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();
-}
-#endif
-
-/*
- * Find the IRQ entry number of a certain pin.
- */
-static int find_irq_entry(int apic, int pin, int type)
-{
-	int i;
-
-	for (i = 0; i < mp_irq_entries; i++)
-		if (mp_irqs[i].mp_irqtype == type &&
-		    (mp_irqs[i].mp_dstapic == mp_ioapics[apic].mp_apicid ||
-		     mp_irqs[i].mp_dstapic == MP_APIC_ALL) &&
-		    mp_irqs[i].mp_dstirq == pin)
-			return i;
-
-	return -1;
-}
-
-/*
- * Find the pin to which IRQ[irq] (ISA) is connected
- */
-static int __init find_isa_irq_pin(int irq, int type)
-{
-	int i;
-
-	for (i = 0; i < mp_irq_entries; i++) {
-		int lbus = mp_irqs[i].mp_srcbus;
-
-		if (test_bit(lbus, mp_bus_not_pci) &&
-		    (mp_irqs[i].mp_irqtype == type) &&
-		    (mp_irqs[i].mp_srcbusirq == irq))
-
-			return mp_irqs[i].mp_dstirq;
-	}
-	return -1;
-}
-
-static int __init find_isa_irq_apic(int irq, int type)
-{
-	int i;
-
-	for (i = 0; i < mp_irq_entries; i++) {
-		int lbus = mp_irqs[i].mp_srcbus;
-
-		if (test_bit(lbus, mp_bus_not_pci) &&
-		    (mp_irqs[i].mp_irqtype == type) &&
-		    (mp_irqs[i].mp_srcbusirq == irq))
-			break;
-	}
-	if (i < mp_irq_entries) {
-		int apic;
-		for(apic = 0; apic < nr_ioapics; apic++) {
-			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic)
-				return apic;
-		}
-	}
-
-	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].mp_srcbus;
-
-		for (apic = 0; apic < nr_ioapics; apic++)
-			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic ||
-			    mp_irqs[i].mp_dstapic == MP_APIC_ALL)
-				break;
-
-		if (!test_bit(lbus, mp_bus_not_pci) &&
-		    !mp_irqs[i].mp_irqtype &&
-		    (bus == lbus) &&
-		    (slot == ((mp_irqs[i].mp_srcbusirq >> 2) & 0x1f))) {
-			int irq = pin_2_irq(i,apic,mp_irqs[i].mp_dstirq);
-
-			if (!(apic || IO_APIC_IRQ(irq)))
-				continue;
-
-			if (pin == (mp_irqs[i].mp_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
- */
-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, KERN_INFO
-			"Broken MPtable reports ISA irq %d\n", irq);
-	return 0;
-}
-
-#endif
-
-/* ISA interrupts are always polarity zero edge triggered,
- * when listed as conforming in the MP table. */
-
-#define default_ISA_trigger(idx)	(0)
-#define default_ISA_polarity(idx)	(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].mp_srcbusirq))
-#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
-
-/* PCI interrupts are always polarity one level triggered,
- * when listed as conforming in the MP table. */
-
-#define default_PCI_trigger(idx)	(1)
-#define default_PCI_polarity(idx)	(1)
-
-/* 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].mp_srcbusirq))
-#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
-
-
-static int MPBIOS_polarity(int idx)
-{
-	int bus = mp_irqs[idx].mp_srcbus;
-	int polarity;
-
-	/*
-	 * Determine IRQ line polarity (high active or low active):
-	 */
-	switch (mp_irqs[idx].mp_irqflag & 3)
-	{
-		case 0: /* conforms, ie. bus-type dependent polarity */
-			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;
-			break;
-		}
-		case 2: /* reserved */
-		{
-			printk(KERN_WARNING "broken BIOS!!\n");
-			polarity = 1;
-			break;
-		}
-		case 3: /* low active */
-		{
-			polarity = 1;
-			break;
-		}
-		default: /* invalid */
-		{
-			printk(KERN_WARNING "broken BIOS!!\n");
-			polarity = 1;
-			break;
-		}
-	}
-	return polarity;
-}
-
-static int MPBIOS_trigger(int idx)
-{
-	int bus = mp_irqs[idx].mp_srcbus;
-	int trigger;
-
-	/*
-	 * Determine IRQ trigger mode (edge or level sensitive):
-	 */
-	switch ((mp_irqs[idx].mp_irqflag>>2) & 3)
-	{
-		case 0: /* conforms, ie. bus-type dependent */
-			if (test_bit(bus, mp_bus_not_pci))
-				trigger = default_ISA_trigger(idx);
-			else
-				trigger = default_PCI_trigger(idx);
-#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
-			switch (mp_bus_id_to_type[bus]) {
-				case MP_BUS_ISA: /* ISA pin */
-				{
-					/* set before the switch */
-					break;
-				}
-				case MP_BUS_EISA: /* EISA pin */
-				{
-					trigger = default_EISA_trigger(idx);
-					break;
-				}
-				case MP_BUS_PCI: /* PCI pin */
-				{
-					/* set before the switch */
-					break;
-				}
-				case MP_BUS_MCA: /* MCA pin */
-				{
-					trigger = default_MCA_trigger(idx);
-					break;
-				}
-				default:
-				{
-					printk(KERN_WARNING "broken BIOS!!\n");
-					trigger = 1;
-					break;
-				}
-			}
-#endif
-			break;
-		case 1: /* edge */
-		{
-			trigger = 0;
-			break;
-		}
-		case 2: /* reserved */
-		{
-			printk(KERN_WARNING "broken BIOS!!\n");
-			trigger = 1;
-			break;
-		}
-		case 3: /* level */
-		{
-			trigger = 1;
-			break;
-		}
-		default: /* invalid */
-		{
-			printk(KERN_WARNING "broken BIOS!!\n");
-			trigger = 0;
-			break;
-		}
-	}
-	return trigger;
-}
-
-static inline int irq_polarity(int idx)
-{
-	return MPBIOS_polarity(idx);
-}
-
-static inline int irq_trigger(int idx)
-{
-	return MPBIOS_trigger(idx);
-}
-
-int (*ioapic_renumber_irq)(int ioapic, int irq);
-static int pin_2_irq(int idx, int apic, int pin)
-{
-	int irq, i;
-	int bus = mp_irqs[idx].mp_srcbus;
-
-	/*
-	 * Debugging check, we are in big trouble if this message pops up!
-	 */
-	if (mp_irqs[idx].mp_dstirq != pin)
-		printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
-
-	if (test_bit(bus, mp_bus_not_pci)) {
-		irq = mp_irqs[idx].mp_srcbusirq;
-	} else {
-		/*
-		 * PCI IRQs are mapped in order
-		 */
-		i = irq = 0;
-		while (i < apic)
-			irq += nr_ioapic_registers[i++];
-		irq += pin;
-                /*
-                 * For MPS mode, so far only needed by ES7000 platform
-                 */
-                if (ioapic_renumber_irq)
-                        irq = ioapic_renumber_irq(apic, irq);
-	}
-
-#ifdef CONFIG_X86_32
-	/*
-	 * PCI IRQ command line redirection. Yes, limits are hardcoded.
-	 */
-	if ((pin >= 16) && (pin <= 23)) {
-		if (pirq_entries[pin-16] != -1) {
-			if (!pirq_entries[pin-16]) {
-				apic_printk(APIC_VERBOSE, KERN_DEBUG
-						"disabling PIRQ%d\n", pin-16);
-			} else {
-				irq = pirq_entries[pin-16];
-				apic_printk(APIC_VERBOSE, KERN_DEBUG
-						"using PIRQ%d -> IRQ %d\n",
-						pin-16, irq);
-			}
-		}
-	}
-#endif
-
-	return irq;
-}
-
-void lock_vector_lock(void)
-{
-	/* Used to the online set of cpus does not change
-	 * during assign_irq_vector.
-	 */
-	spin_lock(&vector_lock);
-}
-
-void unlock_vector_lock(void)
-{
-	spin_unlock(&vector_lock);
-}
-
-static int __assign_irq_vector(int irq, cpumask_t mask)
-{
-	/*
-	 * NOTE! The local APIC isn't very good at handling
-	 * multiple interrupts at the same interrupt level.
-	 * As the interrupt level is determined by taking the
-	 * vector number and shifting that right by 4, we
-	 * want to spread these out a bit so that they don't
-	 * all fall in the same interrupt level.
-	 *
-	 * Also, we've got to be careful not to trash gate
-	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
-	 */
-	static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0;
-	unsigned int old_vector;
-	int cpu;
-	struct irq_cfg *cfg;
-
-	cfg = irq_cfg(irq);
-
-	/* Only try and allocate irqs on cpus that are present */
-	cpus_and(mask, mask, cpu_online_map);
-
-	if ((cfg->move_in_progress) || cfg->move_cleanup_count)
-		return -EBUSY;
-
-	old_vector = cfg->vector;
-	if (old_vector) {
-		cpumask_t tmp;
-		cpus_and(tmp, cfg->domain, mask);
-		if (!cpus_empty(tmp))
-			return 0;
-	}
-
-	for_each_cpu_mask_nr(cpu, mask) {
-		cpumask_t domain, new_mask;
-		int new_cpu;
-		int vector, offset;
-
-		domain = vector_allocation_domain(cpu);
-		cpus_and(new_mask, domain, cpu_online_map);
-
-		vector = current_vector;
-		offset = current_offset;
-next:
-		vector += 8;
-		if (vector >= first_system_vector) {
-			/* If we run out of vectors on large boxen, must share them. */
-			offset = (offset + 1) % 8;
-			vector = FIRST_DEVICE_VECTOR + offset;
-		}
-		if (unlikely(current_vector == vector))
-			continue;
-#ifdef CONFIG_X86_64
-		if (vector == IA32_SYSCALL_VECTOR)
-			goto next;
-#else
-		if (vector == SYSCALL_VECTOR)
-			goto next;
-#endif
-		for_each_cpu_mask_nr(new_cpu, new_mask)
-			if (per_cpu(vector_irq, new_cpu)[vector] != -1)
-				goto next;
-		/* Found one! */
-		current_vector = vector;
-		current_offset = offset;
-		if (old_vector) {
-			cfg->move_in_progress = 1;
-			cfg->old_domain = cfg->domain;
-		}
-		for_each_cpu_mask_nr(new_cpu, new_mask)
-			per_cpu(vector_irq, new_cpu)[vector] = irq;
-		cfg->vector = vector;
-		cfg->domain = domain;
-		return 0;
-	}
-	return -ENOSPC;
-}
-
-static int assign_irq_vector(int irq, cpumask_t mask)
-{
-	int err;
-	unsigned long flags;
-
-	spin_lock_irqsave(&vector_lock, flags);
-	err = __assign_irq_vector(irq, mask);
-	spin_unlock_irqrestore(&vector_lock, flags);
-	return err;
-}
-
-static void __clear_irq_vector(int irq)
-{
-	struct irq_cfg *cfg;
-	cpumask_t mask;
-	int cpu, vector;
-
-	cfg = irq_cfg(irq);
-	BUG_ON(!cfg->vector);
-
-	vector = cfg->vector;
-	cpus_and(mask, cfg->domain, cpu_online_map);
-	for_each_cpu_mask_nr(cpu, mask)
-		per_cpu(vector_irq, cpu)[vector] = -1;
-
-	cfg->vector = 0;
-	cpus_clear(cfg->domain);
-}
-
-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;
-	struct irq_cfg *cfg;
-
-	/* Mark the inuse vectors */
-	for_each_irq_cfg(cfg) {
-		if (!cpu_isset(cpu, cfg->domain))
-			continue;
-		vector = cfg->vector;
-		irq = cfg->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;
-
-		cfg = irq_cfg(irq);
-		if (!cpu_isset(cpu, cfg->domain))
-			per_cpu(vector_irq, cpu)[vector] = -1;
-	}
-}
-
-static struct irq_chip ioapic_chip;
-#ifdef CONFIG_INTR_REMAP
-static struct irq_chip ir_ioapic_chip;
-#endif
-
-#define IOAPIC_AUTO     -1
-#define IOAPIC_EDGE     0
-#define IOAPIC_LEVEL    1
-
-#ifdef CONFIG_X86_32
-static inline int IO_APIC_irq_trigger(int irq)
-{
-        int apic, idx, pin;
-
-        for (apic = 0; apic < nr_ioapics; apic++) {
-                for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-                        idx = find_irq_entry(apic, pin, mp_INT);
-                        if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
-                                return irq_trigger(idx);
-                }
-        }
-        /*
-         * nonexistent IRQs are edge default
-         */
-        return 0;
-}
-#else
-static inline int IO_APIC_irq_trigger(int irq)
-{
-	return 1;
-}
-#endif
-
-static void ioapic_register_intr(int irq, unsigned long trigger)
-{
-	struct irq_desc *desc;
-
-	/* first time to use this irq_desc */
-	if (irq < 16)
-		desc = irq_desc(irq);
-	else
-		desc = irq_desc_with_new(irq);
-
-	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-	    trigger == IOAPIC_LEVEL)
-		desc->status |= IRQ_LEVEL;
-	else
-		desc->status &= ~IRQ_LEVEL;
-
-#ifdef CONFIG_INTR_REMAP
-	if (irq_remapped(irq)) {
-		desc->status |= IRQ_MOVE_PCNTXT;
-		if (trigger)
-			set_irq_chip_and_handler_name(irq, &ir_ioapic_chip,
-						      handle_fasteoi_irq,
-						     "fasteoi");
-		else
-			set_irq_chip_and_handler_name(irq, &ir_ioapic_chip,
-						      handle_edge_irq, "edge");
-		return;
-	}
-#endif
-	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
-	    trigger == IOAPIC_LEVEL)
-		set_irq_chip_and_handler_name(irq, &ioapic_chip,
-					      handle_fasteoi_irq,
-					      "fasteoi");
-	else
-		set_irq_chip_and_handler_name(irq, &ioapic_chip,
-					      handle_edge_irq, "edge");
-}
-
-static int setup_ioapic_entry(int apic, int irq,
-			      struct IO_APIC_route_entry *entry,
-			      unsigned int destination, int trigger,
-			      int polarity, int vector)
-{
-	/*
-	 * add it to the IO-APIC irq-routing table:
-	 */
-	memset(entry,0,sizeof(*entry));
-
-#ifdef CONFIG_INTR_REMAP
-	if (intr_remapping_enabled) {
-		struct intel_iommu *iommu = map_ioapic_to_ir(apic);
-		struct irte irte;
-		struct IR_IO_APIC_route_entry *ir_entry =
-			(struct IR_IO_APIC_route_entry *) entry;
-		int index;
-
-		if (!iommu)
-			panic("No mapping iommu for ioapic %d\n", apic);
-
-		index = alloc_irte(iommu, irq, 1);
-		if (index < 0)
-			panic("Failed to allocate IRTE for ioapic %d\n", apic);
-
-		memset(&irte, 0, sizeof(irte));
-
-		irte.present = 1;
-		irte.dst_mode = INT_DEST_MODE;
-		irte.trigger_mode = trigger;
-		irte.dlvry_mode = INT_DELIVERY_MODE;
-		irte.vector = vector;
-		irte.dest_id = IRTE_DEST(destination);
-
-		modify_irte(irq, &irte);
-
-		ir_entry->index2 = (index >> 15) & 0x1;
-		ir_entry->zero = 0;
-		ir_entry->format = 1;
-		ir_entry->index = (index & 0x7fff);
-	} else
-#endif
-	{
-		entry->delivery_mode = INT_DELIVERY_MODE;
-		entry->dest_mode = INT_DEST_MODE;
-		entry->dest = destination;
-	}
-
-	entry->mask = 0;				/* enable IRQ */
-	entry->trigger = trigger;
-	entry->polarity = polarity;
-	entry->vector = vector;
-
-	/* Mask level triggered irqs.
-	 * Use IRQ_DELAYED_DISABLE for edge triggered irqs.
-	 */
-	if (trigger)
-		entry->mask = 1;
-	return 0;
-}
-
-static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq,
-			      int trigger, int polarity)
-{
-	struct irq_cfg *cfg;
-	struct IO_APIC_route_entry entry;
-	cpumask_t mask;
-
-	if (!IO_APIC_IRQ(irq))
-		return;
-
-	cfg = irq_cfg(irq);
-
-	mask = TARGET_CPUS;
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cpus_and(mask, cfg->domain, mask);
-
-	apic_printk(APIC_VERBOSE,KERN_DEBUG
-		    "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> "
-		    "IRQ %d Mode:%i Active:%i)\n",
-		    apic, mp_ioapics[apic].mp_apicid, pin, cfg->vector,
-		    irq, trigger, polarity);
-
-
-	if (setup_ioapic_entry(mp_ioapics[apic].mp_apicid, irq, &entry,
-			       cpu_mask_to_apicid(mask), trigger, polarity,
-			       cfg->vector)) {
-		printk("Failed to setup ioapic entry for ioapic  %d, pin %d\n",
-		       mp_ioapics[apic].mp_apicid, pin);
-		__clear_irq_vector(irq);
-		return;
-	}
-
-	ioapic_register_intr(irq, trigger);
-	if (irq < 16)
-		disable_8259A_irq(irq);
-
-	ioapic_write_entry(apic, pin, entry);
-}
-
-static void __init setup_IO_APIC_irqs(void)
-{
-	int apic, pin, idx, irq, first_notcon = 1;
-
-	apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-	for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-
-		idx = find_irq_entry(apic,pin,mp_INT);
-		if (idx == -1) {
-			if (first_notcon) {
-				apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mp_apicid, pin);
-				first_notcon = 0;
-			} else
-				apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mp_apicid, pin);
-			continue;
-		}
-		if (!first_notcon) {
-			apic_printk(APIC_VERBOSE, " not connected.\n");
-			first_notcon = 1;
-		}
-
-		irq = pin_2_irq(idx, apic, pin);
-#ifdef CONFIG_X86_32
-                if (multi_timer_check(apic, irq))
-                        continue;
-#endif
-		add_pin_to_irq(irq, apic, pin);
-
-		setup_IO_APIC_irq(apic, pin, irq,
-				  irq_trigger(idx), irq_polarity(idx));
-	}
-	}
-
-	if (!first_notcon)
-		apic_printk(APIC_VERBOSE, " not connected.\n");
-}
-
-/*
- * Set up the timer pin, possibly with the 8259A-master behind.
- */
-static void __init setup_timer_IRQ0_pin(unsigned int apic, unsigned int pin,
-					int vector)
-{
-	struct IO_APIC_route_entry entry;
-
-#ifdef CONFIG_INTR_REMAP
-	if (intr_remapping_enabled)
-		return;
-#endif
-
-	memset(&entry, 0, sizeof(entry));
-
-	/*
-	 * We use logical delivery to get the timer IRQ
-	 * to the first CPU.
-	 */
-	entry.dest_mode = INT_DEST_MODE;
-	entry.mask = 1;					/* mask IRQ now */
-	entry.dest = cpu_mask_to_apicid(TARGET_CPUS);
-	entry.delivery_mode = INT_DELIVERY_MODE;
-	entry.polarity = 0;
-	entry.trigger = 0;
-	entry.vector = vector;
-
-	/*
-	 * The timer IRQ doesn't have to know that behind the
-	 * scene we may have a 8259A-master in AEOI mode ...
-	 */
-	set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge");
-
-	/*
-	 * Add it to the IO-APIC irq-routing table:
-	 */
-	ioapic_write_entry(apic, pin, entry);
-}
-
-
-__apicdebuginit(void) print_IO_APIC(void)
-{
-	int apic, i;
-	union IO_APIC_reg_00 reg_00;
-	union IO_APIC_reg_01 reg_01;
-	union IO_APIC_reg_02 reg_02;
-	union IO_APIC_reg_03 reg_03;
-	unsigned long flags;
-	struct irq_cfg *cfg;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries);
-	for (i = 0; i < nr_ioapics; i++)
-		printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n",
-		       mp_ioapics[i].mp_apicid, nr_ioapic_registers[i]);
-
-	/*
-	 * We are a bit conservative about what we expect.  We have to
-	 * know about every hardware change ASAP.
-	 */
-	printk(KERN_INFO "testing the IO APIC.......................\n");
-
-	for (apic = 0; apic < nr_ioapics; apic++) {
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	reg_00.raw = io_apic_read(apic, 0);
-	reg_01.raw = io_apic_read(apic, 1);
-	if (reg_01.bits.version >= 0x10)
-		reg_02.raw = io_apic_read(apic, 2);
-        if (reg_01.bits.version >= 0x20)
-                reg_03.raw = io_apic_read(apic, 3);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	printk("\n");
-	printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid);
-	printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
-	printk(KERN_DEBUG ".......    : physical APIC id: %02X\n", reg_00.bits.ID);
-	printk(KERN_DEBUG ".......    : Delivery Type: %X\n", reg_00.bits.delivery_type);
-	printk(KERN_DEBUG ".......    : LTS          : %X\n", reg_00.bits.LTS);
-
-	printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)&reg_01);
-	printk(KERN_DEBUG ".......     : max redirection entries: %04X\n", reg_01.bits.entries);
-
-	printk(KERN_DEBUG ".......     : PRQ implemented: %X\n", reg_01.bits.PRQ);
-	printk(KERN_DEBUG ".......     : IO APIC version: %04X\n", reg_01.bits.version);
-
-	/*
-	 * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02,
-	 * but the value of reg_02 is read as the previous read register
-	 * value, so ignore it if reg_02 == reg_01.
-	 */
-	if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) {
-		printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
-		printk(KERN_DEBUG ".......     : arbitration: %02X\n", reg_02.bits.arbitration);
-	}
-
-	/*
-	 * Some Intel chipsets with IO APIC VERSION of 0x2? don't have reg_02
-	 * or reg_03, but the value of reg_0[23] is read as the previous read
-	 * register value, so ignore it if reg_03 == reg_0[12].
-	 */
-	if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw &&
-	    reg_03.raw != reg_01.raw) {
-		printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw);
-		printk(KERN_DEBUG ".......     : Boot DT    : %X\n", reg_03.bits.boot_DT);
-	}
-
-	printk(KERN_DEBUG ".... IRQ redirection table:\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;
-
-		entry = ioapic_read_entry(apic, i);
-
-		printk(KERN_DEBUG " %02x %03X ",
-			i,
-			entry.dest
-		);
-
-		printk("%1d    %1d    %1d   %1d   %1d    %1d    %1d    %02X\n",
-			entry.mask,
-			entry.trigger,
-			entry.irr,
-			entry.polarity,
-			entry.delivery_status,
-			entry.dest_mode,
-			entry.delivery_mode,
-			entry.vector
-		);
-	}
-	}
-	printk(KERN_DEBUG "IRQ to pin mappings:\n");
-	for_each_irq_cfg(cfg) {
-		struct irq_pin_list *entry = cfg->irq_2_pin;
-		if (!entry)
-			continue;
-		printk(KERN_DEBUG "IRQ%d ", cfg->irq);
-		for (;;) {
-			printk("-> %d:%d", entry->apic, entry->pin);
-			if (!entry->next)
-				break;
-			entry = entry->next;
-		}
-		printk("\n");
-	}
-
-	printk(KERN_INFO ".................................... done.\n");
-
-	return;
-}
-
-__apicdebuginit(void) print_APIC_bitfield(int base)
-{
-	unsigned int v;
-	int i, j;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
-	for (i = 0; i < 8; i++) {
-		v = apic_read(base + i*0x10);
-		for (j = 0; j < 32; j++) {
-			if (v & (1<<j))
-				printk("1");
-			else
-				printk("0");
-		}
-		printk("\n");
-	}
-}
-
-__apicdebuginit(void) print_local_APIC(void *dummy)
-{
-	unsigned int v, ver, maxlvt;
-	u64 icr;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
-		smp_processor_id(), hard_smp_processor_id());
-	v = apic_read(APIC_ID);
-	printk(KERN_INFO "... APIC ID:      %08x (%01x)\n", v, read_apic_id());
-	v = apic_read(APIC_LVR);
-	printk(KERN_INFO "... APIC VERSION: %08x\n", v);
-	ver = GET_APIC_VERSION(v);
-	maxlvt = lapic_get_maxlvt();
-
-	v = apic_read(APIC_TASKPRI);
-	printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
-
-	if (APIC_INTEGRATED(ver)) {                     /* !82489DX */
-		v = apic_read(APIC_ARBPRI);
-		printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
-			v & APIC_ARBPRI_MASK);
-		v = apic_read(APIC_PROCPRI);
-		printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
-	}
-
-	v = apic_read(APIC_EOI);
-	printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
-	v = apic_read(APIC_RRR);
-	printk(KERN_DEBUG "... APIC RRR: %08x\n", v);
-	v = apic_read(APIC_LDR);
-	printk(KERN_DEBUG "... APIC LDR: %08x\n", v);
-	v = apic_read(APIC_DFR);
-	printk(KERN_DEBUG "... APIC DFR: %08x\n", v);
-	v = apic_read(APIC_SPIV);
-	printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
-
-	printk(KERN_DEBUG "... APIC ISR field:\n");
-	print_APIC_bitfield(APIC_ISR);
-	printk(KERN_DEBUG "... APIC TMR field:\n");
-	print_APIC_bitfield(APIC_TMR);
-	printk(KERN_DEBUG "... APIC IRR field:\n");
-	print_APIC_bitfield(APIC_IRR);
-
-	if (APIC_INTEGRATED(ver)) {             /* !82489DX */
-		if (maxlvt > 3)         /* Due to the Pentium erratum 3AP. */
-			apic_write(APIC_ESR, 0);
-
-		v = apic_read(APIC_ESR);
-		printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
-	}
-
-	icr = apic_icr_read();
-	printk(KERN_DEBUG "... APIC ICR: %08x\n", icr);
-	printk(KERN_DEBUG "... APIC ICR2: %08x\n", icr >> 32);
-
-	v = apic_read(APIC_LVTT);
-	printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
-
-	if (maxlvt > 3) {                       /* PC is LVT#4. */
-		v = apic_read(APIC_LVTPC);
-		printk(KERN_DEBUG "... APIC LVTPC: %08x\n", v);
-	}
-	v = apic_read(APIC_LVT0);
-	printk(KERN_DEBUG "... APIC LVT0: %08x\n", v);
-	v = apic_read(APIC_LVT1);
-	printk(KERN_DEBUG "... APIC LVT1: %08x\n", v);
-
-	if (maxlvt > 2) {			/* ERR is LVT#3. */
-		v = apic_read(APIC_LVTERR);
-		printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
-	}
-
-	v = apic_read(APIC_TMICT);
-	printk(KERN_DEBUG "... APIC TMICT: %08x\n", v);
-	v = apic_read(APIC_TMCCT);
-	printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
-	v = apic_read(APIC_TDCR);
-	printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
-	printk("\n");
-}
-
-__apicdebuginit(void) print_all_local_APICs(void)
-{
-	on_each_cpu(print_local_APIC, NULL, 1);
-}
-
-__apicdebuginit(void) print_PIC(void)
-{
-	unsigned int v;
-	unsigned long flags;
-
-	if (apic_verbosity == APIC_QUIET)
-		return;
-
-	printk(KERN_DEBUG "\nprinting PIC contents\n");
-
-	spin_lock_irqsave(&i8259A_lock, flags);
-
-	v = inb(0xa1) << 8 | inb(0x21);
-	printk(KERN_DEBUG "... PIC  IMR: %04x\n", v);
-
-	v = inb(0xa0) << 8 | inb(0x20);
-	printk(KERN_DEBUG "... PIC  IRR: %04x\n", v);
-
-	outb(0x0b,0xa0);
-	outb(0x0b,0x20);
-	v = inb(0xa0) << 8 | inb(0x20);
-	outb(0x0a,0xa0);
-	outb(0x0a,0x20);
-
-	spin_unlock_irqrestore(&i8259A_lock, flags);
-
-	printk(KERN_DEBUG "... PIC  ISR: %04x\n", v);
-
-	v = inb(0x4d1) << 8 | inb(0x4d0);
-	printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
-}
-
-__apicdebuginit(int) print_all_ICs(void)
-{
-	print_PIC();
-	print_all_local_APICs();
-	print_IO_APIC();
-
-	return 0;
-}
-
-fs_initcall(print_all_ICs);
-
-
-/* Where if anywhere is the i8259 connect in external int mode */
-static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
-
-void __init enable_IO_APIC(void)
-{
-	union IO_APIC_reg_01 reg_01;
-	int i8259_apic, i8259_pin;
-	int apic;
-	unsigned long flags;
-
-#ifdef CONFIG_X86_32
-	int i;
-	if (!pirqs_enabled)
-		for (i = 0; i < MAX_PIRQS; i++)
-			pirq_entries[i] = -1;
-#endif
-
-	/*
-	 * The number of IO-APIC IRQ registers (== #pins):
-	 */
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		spin_lock_irqsave(&ioapic_lock, flags);
-		reg_01.raw = io_apic_read(apic, 1);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
-	}
-	for(apic = 0; apic < nr_ioapics; apic++) {
-		int pin;
-		/* See if any of the pins is in ExtINT mode */
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-			struct IO_APIC_route_entry entry;
-			entry = ioapic_read_entry(apic, pin);
-
-			/* If the interrupt line is enabled and in ExtInt mode
-			 * I have found the pin where the i8259 is connected.
-			 */
-			if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) {
-				ioapic_i8259.apic = apic;
-				ioapic_i8259.pin  = pin;
-				goto found_i8259;
-			}
-		}
-	}
- found_i8259:
-	/* Look to see what if the MP table has reported the ExtINT */
-	/* If we could not find the appropriate pin by looking at the ioapic
-	 * the i8259 probably is not connected the ioapic but give the
-	 * mptable a chance anyway.
-	 */
-	i8259_pin  = find_isa_irq_pin(0, mp_ExtINT);
-	i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
-	/* Trust the MP table if nothing is setup in the hardware */
-	if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) {
-		printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n");
-		ioapic_i8259.pin  = i8259_pin;
-		ioapic_i8259.apic = i8259_apic;
-	}
-	/* Complain if the MP table and the hardware disagree */
-	if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) &&
-		(i8259_pin >= 0) && (ioapic_i8259.pin >= 0))
-	{
-		printk(KERN_WARNING "ExtINT in hardware and MP table differ\n");
-	}
-
-	/*
-	 * Do not trust the IO-APIC being empty at bootup
-	 */
-	clear_IO_APIC();
-}
-
-/*
- * Not an __init, needed by the reboot code
- */
-void disable_IO_APIC(void)
-{
-	/*
-	 * Clear the IO-APIC before rebooting:
-	 */
-	clear_IO_APIC();
-
-	/*
-	 * 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;
-
-		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            = read_apic_id();
-
-		/*
-		 * Add it to the IO-APIC irq-routing table:
-		 */
-		ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry);
-	}
-
-	disconnect_bsp_APIC(ioapic_i8259.pin != -1);
-}
-
-#ifdef CONFIG_X86_32
-/*
- * function to set the IO-APIC physical IDs based on the
- * values stored in the MPC table.
- *
- * by Matt Domsch <Matt_Domsch@dell.com>  Tue Dec 21 12:25:05 CST 1999
- */
-
-static void __init setup_ioapic_ids_from_mpc(void)
-{
-	union IO_APIC_reg_00 reg_00;
-	physid_mask_t phys_id_present_map;
-	int apic;
-	int i;
-	unsigned char old_id;
-	unsigned long flags;
-
-	if (x86_quirks->setup_ioapic_ids && x86_quirks->setup_ioapic_ids())
-		return;
-
-	/*
-	 * Don't check I/O APIC IDs for xAPIC systems.  They have
-	 * no meaning without the serial APIC bus.
-	 */
-	if (!(boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
-		|| APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
-		return;
-	/*
-	 * This is broken; anything with a real cpu count has to
-	 * circumvent this idiocy regardless.
-	 */
-	phys_id_present_map = ioapic_phys_id_map(phys_cpu_present_map);
-
-	/*
-	 * 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].mp_apicid;
-
-		if (mp_ioapics[apic].mp_apicid >= get_physical_broadcast()) {
-			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n",
-				apic, mp_ioapics[apic].mp_apicid);
-			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
-				reg_00.bits.ID);
-			mp_ioapics[apic].mp_apicid = reg_00.bits.ID;
-		}
-
-		/*
-		 * Sanity check, is the ID really free? Every APIC in a
-		 * system must have a unique ID or we get lots of nice
-		 * 'stuck on smp_invalidate_needed IPI wait' messages.
-		 */
-		if (check_apicid_used(phys_id_present_map,
-					mp_ioapics[apic].mp_apicid)) {
-			printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n",
-				apic, mp_ioapics[apic].mp_apicid);
-			for (i = 0; i < get_physical_broadcast(); i++)
-				if (!physid_isset(i, phys_id_present_map))
-					break;
-			if (i >= get_physical_broadcast())
-				panic("Max APIC ID exceeded!\n");
-			printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n",
-				i);
-			physid_set(i, phys_id_present_map);
-			mp_ioapics[apic].mp_apicid = i;
-		} else {
-			physid_mask_t tmp;
-			tmp = apicid_to_cpu_present(mp_ioapics[apic].mp_apicid);
-			apic_printk(APIC_VERBOSE, "Setting %d in the "
-					"phys_id_present_map\n",
-					mp_ioapics[apic].mp_apicid);
-			physids_or(phys_id_present_map, phys_id_present_map, tmp);
-		}
-
-
-		/*
-		 * We need to adjust the IRQ routing table
-		 * if the ID changed.
-		 */
-		if (old_id != mp_ioapics[apic].mp_apicid)
-			for (i = 0; i < mp_irq_entries; i++)
-				if (mp_irqs[i].mp_dstapic == old_id)
-					mp_irqs[i].mp_dstapic
-						= mp_ioapics[apic].mp_apicid;
-
-		/*
-		 * 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].mp_apicid);
-
-		reg_00.bits.ID = mp_ioapics[apic].mp_apicid;
-		spin_lock_irqsave(&ioapic_lock, flags);
-
-		/*
-		 * Sanity check
-		 */
-		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].mp_apicid)
-			printk("could not set ID!\n");
-		else
-			apic_printk(APIC_VERBOSE, " ok.\n");
-	}
-}
-#endif
-
-int no_timer_check __initdata;
-
-static int __init notimercheck(char *s)
-{
-	no_timer_check = 1;
-	return 1;
-}
-__setup("no_timer_check", notimercheck);
-
-/*
- * There is a nasty bug in some older SMP boards, their mptable lies
- * about the timer IRQ. We do the following to work around the situation:
- *
- *	- timer IRQ defaults to IO-APIC IRQ
- *	- if this function detects that timer IRQs are defunct, then we fall
- *	  back to ISA timer IRQs
- */
-static int __init timer_irq_works(void)
-{
-	unsigned long t1 = jiffies;
-	unsigned long flags;
-
-	if (no_timer_check)
-		return 1;
-
-	local_save_flags(flags);
-	local_irq_enable();
-	/* Let ten ticks pass... */
-	mdelay((10 * 1000) / HZ);
-	local_irq_restore(flags);
-
-	/*
-	 * Expect a few ticks at least, to be sure some possible
-	 * glue logic does not lock up after one or two first
-	 * ticks in a non-ExtINT mode.  Also the local APIC
-	 * might have cached one ExtINT interrupt.  Finally, at
-	 * least one tick may be lost due to delays.
-	 */
-
-	/* jiffies wrap? */
-	if (time_after(jiffies, t1 + 4))
-		return 1;
-	return 0;
-}
-
-/*
- * In the SMP+IOAPIC case it might happen that there are an unspecified
- * number of pending IRQ events unhandled. These cases are very rare,
- * so we 'resend' these IRQs via IPIs, to the same CPU. It's much
- * better to do it this way as thus we do not have to be aware of
- * 'pending' interrupts in the IRQ path, except at this point.
- */
-/*
- * Edge triggered needs to resend any interrupt
- * that was delayed but this is now handled in the device
- * independent code.
- */
-
-/*
- * Starting up a edge-triggered IO-APIC interrupt is
- * nasty - we need to make sure that we get the edge.
- * If it is already asserted for some reason, we need
- * return 1 to indicate that is was pending.
- *
- * This is not complete - we should be able to fake
- * an edge even if it isn't on the 8259A...
- */
-
-static unsigned int startup_ioapic_irq(unsigned int irq)
-{
-	int was_pending = 0;
-	unsigned long flags;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	if (irq < 16) {
-		disable_8259A_irq(irq);
-		if (i8259A_irq_pending(irq))
-			was_pending = 1;
-	}
-	__unmask_IO_APIC_irq(irq);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	return was_pending;
-}
-
-#ifdef CONFIG_X86_64
-static int ioapic_retrigger_irq(unsigned int irq)
-{
-
-	struct irq_cfg *cfg = irq_cfg(irq);
-	unsigned long flags;
-
-	spin_lock_irqsave(&vector_lock, flags);
-	send_IPI_mask(cpumask_of_cpu(first_cpu(cfg->domain)), cfg->vector);
-	spin_unlock_irqrestore(&vector_lock, flags);
-
-	return 1;
-}
-#else
-static int ioapic_retrigger_irq(unsigned int irq)
-{
-        send_IPI_self(irq_cfg(irq)->vector);
-
-        return 1;
-}
-#endif
-
-/*
- * Level and edge triggered IO-APIC interrupts need different handling,
- * so we use two separate IRQ descriptors. Edge triggered IRQs can be
- * handled with the level-triggered descriptor, but that one has slightly
- * more overhead. Level-triggered interrupts cannot be handled with the
- * edge-triggered handler, without risking IRQ storms and other ugly
- * races.
- */
-
-#ifdef CONFIG_SMP
-
-#ifdef CONFIG_INTR_REMAP
-static void ir_irq_migration(struct work_struct *work);
-
-static DECLARE_DELAYED_WORK(ir_migration_work, ir_irq_migration);
-
-/*
- * Migrate the IO-APIC irq in the presence of intr-remapping.
- *
- * For edge triggered, irq migration is a simple atomic update(of vector
- * and cpu destination) of IRTE and flush the hardware cache.
- *
- * For level triggered, we need to modify the io-apic RTE aswell with the update
- * vector information, along with modifying IRTE with vector and destination.
- * So irq migration for level triggered is little  bit more complex compared to
- * edge triggered migration. But the good news is, we use the same algorithm
- * for level triggered migration as we have today, only difference being,
- * we now initiate the irq migration from process context instead of the
- * interrupt context.
- *
- * In future, when we do a directed EOI (combined with cpu EOI broadcast
- * suppression) to the IO-APIC, level triggered irq migration will also be
- * as simple as edge triggered migration and we can do the irq migration
- * with a simple atomic update to IO-APIC RTE.
- */
-static void migrate_ioapic_irq(int irq, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	struct irq_desc *desc;
-	cpumask_t tmp, cleanup_mask;
-	struct irte irte;
-	int modify_ioapic_rte;
-	unsigned int dest;
-	unsigned long flags;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (get_irte(irq, &irte))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	desc = irq_desc(irq);
-	modify_ioapic_rte = desc->status & IRQ_LEVEL;
-	if (modify_ioapic_rte) {
-		spin_lock_irqsave(&ioapic_lock, flags);
-		__target_IO_APIC_irq(irq, dest, cfg->vector);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-	}
-
-	irte.vector = cfg->vector;
-	irte.dest_id = IRTE_DEST(dest);
-
-	/*
-	 * Modified the IRTE and flushes the Interrupt entry cache.
-	 */
-	modify_irte(irq, &irte);
-
-	if (cfg->move_in_progress) {
-		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
-		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
-		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
-		cfg->move_in_progress = 0;
-	}
-
-	desc->affinity = mask;
-}
-
-static int migrate_irq_remapped_level(int irq)
-{
-	int ret = -1;
-	struct irq_desc *desc = irq_desc(irq);
-
-	mask_IO_APIC_irq(irq);
-
-	if (io_apic_level_ack_pending(irq)) {
-		/*
-	 	 * Interrupt in progress. Migrating irq now will change the
-		 * vector information in the IO-APIC RTE and that will confuse
-		 * the EOI broadcast performed by cpu.
-		 * So, delay the irq migration to the next instance.
-		 */
-		schedule_delayed_work(&ir_migration_work, 1);
-		goto unmask;
-	}
-
-	/* everthing is clear. we have right of way */
-	migrate_ioapic_irq(irq, desc->pending_mask);
-
-	ret = 0;
-	desc->status &= ~IRQ_MOVE_PENDING;
-	cpus_clear(desc->pending_mask);
-
-unmask:
-	unmask_IO_APIC_irq(irq);
-	return ret;
-}
-
-static void ir_irq_migration(struct work_struct *work)
-{
-	unsigned int irq;
-	struct irq_desc *desc;
-
-	for_each_irq_desc(irq, desc) {
-		if (desc->status & IRQ_MOVE_PENDING) {
-			unsigned long flags;
-
-			spin_lock_irqsave(&desc->lock, flags);
-			if (!desc->chip->set_affinity ||
-			    !(desc->status & IRQ_MOVE_PENDING)) {
-				desc->status &= ~IRQ_MOVE_PENDING;
-				spin_unlock_irqrestore(&desc->lock, flags);
-				continue;
-			}
-
-			desc->chip->set_affinity(irq, desc, desc->pending_mask);
-			spin_unlock_irqrestore(&desc->lock, flags);
-		}
-	}
-}
-
-/*
- * Migrates the IRQ destination in the process context.
- */
-static void set_ir_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	if (desc->status & IRQ_LEVEL) {
-		desc->status |= IRQ_MOVE_PENDING;
-		desc->pending_mask = mask;
-		migrate_irq_remapped_level(irq);
-		return;
-	}
-
-	migrate_ioapic_irq(irq, mask);
-}
-#endif
-
-asmlinkage void smp_irq_move_cleanup_interrupt(void)
-{
-	unsigned vector, me;
-	ack_APIC_irq();
-#ifdef CONFIG_X86_64
-	exit_idle();
-#endif
-	irq_enter();
-
-	me = smp_processor_id();
-	for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
-		unsigned int irq;
-		struct irq_desc *desc;
-		struct irq_cfg *cfg;
-		irq = __get_cpu_var(vector_irq)[vector];
-
-		desc = irq_desc(irq);
-		if (!desc)
-			continue;
-
-		cfg = irq_cfg(irq);
-		spin_lock(&desc->lock);
-		if (!cfg->move_cleanup_count)
-			goto unlock;
-
-		if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
-			goto unlock;
-
-		__get_cpu_var(vector_irq)[vector] = -1;
-		cfg->move_cleanup_count--;
-unlock:
-		spin_unlock(&desc->lock);
-	}
-
-	irq_exit();
-}
-
-static void irq_complete_move(unsigned int irq)
-{
-	struct irq_cfg *cfg = irq_cfg(irq);
-	unsigned vector, me;
-
-	if (likely(!cfg->move_in_progress))
-		return;
-
-	vector = ~get_irq_regs()->orig_ax;
-	me = smp_processor_id();
-	if ((vector == cfg->vector) && cpu_isset(me, cfg->domain)) {
-		cpumask_t cleanup_mask;
-
-		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
-		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
-		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
-		cfg->move_in_progress = 0;
-	}
-}
-#else
-static inline void irq_complete_move(unsigned int irq) {}
-#endif
-#ifdef CONFIG_INTR_REMAP
-static void ack_x2apic_level(unsigned int irq, struct irq_desc *desc)
-{
-	ack_x2APIC_irq();
-}
-
-static void ack_x2apic_edge(unsigned int irq, struct irq_desc *desc)
-{
-	ack_x2APIC_irq();
-}
-#endif
-
-static void ack_apic_edge(unsigned int irq, struct irq_desc *desc)
-{
-	irq_complete_move(irq);
-	move_native_irq(irq, desc);
-	ack_APIC_irq();
-}
-
-#ifdef CONFIG_X86_64
-static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
-{
-	int do_unmask_irq = 0;
-
-	irq_complete_move(irq);
-#ifdef CONFIG_GENERIC_PENDING_IRQ
-	/* If we are moving the irq we need to mask it */
-	if (unlikely(desc->status & IRQ_MOVE_PENDING)) {
-		do_unmask_irq = 1;
-		mask_IO_APIC_irq(irq);
-	}
-#endif
-
-	/*
-	 * We must acknowledge the irq before we move it or the acknowledge will
-	 * not propagate properly.
-	 */
-	ack_APIC_irq();
-
-	/* Now we can move and renable the irq */
-	if (unlikely(do_unmask_irq)) {
-		/* Only migrate the irq if the ack has been received.
-		 *
-		 * On rare occasions the broadcast level triggered ack gets
-		 * delayed going to ioapics, and if we reprogram the
-		 * vector while Remote IRR is still set the irq will never
-		 * fire again.
-		 *
-		 * To prevent this scenario we read the Remote IRR bit
-		 * of the ioapic.  This has two effects.
-		 * - On any sane system the read of the ioapic will
-		 *   flush writes (and acks) going to the ioapic from
-		 *   this cpu.
-		 * - We get to see if the ACK has actually been delivered.
-		 *
-		 * Based on failed experiments of reprogramming the
-		 * ioapic entry from outside of irq context starting
-		 * with masking the ioapic entry and then polling until
-		 * Remote IRR was clear before reprogramming the
-		 * ioapic I don't trust the Remote IRR bit to be
-		 * completey accurate.
-		 *
-		 * However there appears to be no other way to plug
-		 * this race, so if the Remote IRR bit is not
-		 * accurate and is causing problems then it is a hardware bug
-		 * and you can go talk to the chipset vendor about it.
-		 */
-		if (!io_apic_level_ack_pending(irq))
-			move_masked_irq(irq, desc);
-		unmask_IO_APIC_irq(irq);
-	}
-}
-#else
-atomic_t irq_mis_count;
-static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
-{
-	unsigned long v;
-	int i;
-
-	irq_complete_move(irq);
-	move_native_irq(irq, desc);
-	/*
-	* 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
-	*/
-	i = irq_cfg(irq)->vector;
-
-	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
-
-	ack_APIC_irq();
-
-	if (!(v & (1 << (i & 0x1f)))) {
-		atomic_inc(&irq_mis_count);
-		spin_lock(&ioapic_lock);
-		__mask_and_edge_IO_APIC_irq(irq);
-		__unmask_and_level_IO_APIC_irq(irq);
-		spin_unlock(&ioapic_lock);
-	}
-}
-#endif
-
-static struct irq_chip ioapic_chip __read_mostly = {
-	.name 		= "IO-APIC",
-	.startup 	= startup_ioapic_irq,
-	.mask	 	= mask_IO_APIC_irq,
-	.unmask	 	= unmask_IO_APIC_irq,
-	.ack 		= ack_apic_edge,
-	.eoi 		= ack_apic_level,
-#ifdef CONFIG_SMP
-	.set_affinity 	= set_ioapic_affinity_irq,
-#endif
-	.retrigger	= ioapic_retrigger_irq,
-};
-
-#ifdef CONFIG_INTR_REMAP
-static struct irq_chip ir_ioapic_chip __read_mostly = {
-	.name 		= "IR-IO-APIC",
-	.startup 	= startup_ioapic_irq,
-	.mask	 	= mask_IO_APIC_irq,
-	.unmask	 	= unmask_IO_APIC_irq,
-	.ack 		= ack_x2apic_edge,
-	.eoi 		= ack_x2apic_level,
-#ifdef CONFIG_SMP
-	.set_affinity 	= set_ir_ioapic_affinity_irq,
-#endif
-	.retrigger	= ioapic_retrigger_irq,
-};
-#endif
-
-static inline void init_IO_APIC_traps(void)
-{
-	int irq;
-	struct irq_desc *desc;
-	struct irq_cfg *cfg;
-
-	/*
-	 * NOTE! The local APIC isn't very good at handling
-	 * multiple interrupts at the same interrupt level.
-	 * As the interrupt level is determined by taking the
-	 * vector number and shifting that right by 4, we
-	 * want to spread these out a bit so that they don't
-	 * all fall in the same interrupt level.
-	 *
-	 * Also, we've got to be careful not to trash gate
-	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
-	 */
-	for_each_irq_cfg(cfg) {
-		irq = cfg->irq;
-		if (IO_APIC_IRQ(irq) && !cfg->vector) {
-			/*
-			 * Hmm.. We don't have an entry for this,
-			 * so default to an old-fashioned 8259
-			 * interrupt if we can..
-			 */
-			if (irq < 16)
-				make_8259A_irq(irq);
-			else {
-				desc = irq_desc(irq);
-				/* Strange. Oh, well.. */
-				desc->chip = &no_irq_chip;
-			}
-		}
-	}
-}
-
-/*
- * The local APIC irq-chip implementation:
- */
-
-static void mask_lapic_irq(unsigned int irq)
-{
-	unsigned long v;
-
-	v = apic_read(APIC_LVT0);
-	apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
-}
-
-static void unmask_lapic_irq(unsigned int irq)
-{
-	unsigned long v;
-
-	v = apic_read(APIC_LVT0);
-	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
-}
-
-static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
-{
-	ack_APIC_irq();
-}
-
-static struct irq_chip lapic_chip __read_mostly = {
-	.name		= "local-APIC",
-	.mask		= mask_lapic_irq,
-	.unmask		= unmask_lapic_irq,
-	.ack		= ack_lapic_irq,
-};
-
-static void lapic_register_intr(int irq)
-{
-	struct irq_desc *desc;
-
-	desc = irq_desc(irq);
-	desc->status &= ~IRQ_LEVEL;
-	set_irq_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq,
-				      "edge");
-}
-
-static void __init setup_nmi(void)
-{
-	/*
-	 * Dirty trick to enable the NMI watchdog ...
-	 * We put the 8259A master into AEOI mode and
-	 * unmask on all local APICs LVT0 as NMI.
-	 *
-	 * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
-	 * is from Maciej W. Rozycki - so we do not have to EOI from
-	 * the NMI handler or the timer interrupt.
-	 */
-	apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ...");
-
-	enable_NMI_through_LVT0();
-
-	apic_printk(APIC_VERBOSE, " done.\n");
-}
-
-/*
- * This looks a bit hackish but it's about the only one way of sending
- * a few INTA cycles to 8259As and any associated glue logic.  ICR does
- * not support the ExtINT mode, unfortunately.  We need to send these
- * cycles as some i82489DX-based boards have glue logic that keeps the
- * 8259A interrupt line asserted until INTA.  --macro
- */
-static inline void __init unlock_ExtINT_logic(void)
-{
-	int apic, pin, i;
-	struct IO_APIC_route_entry entry0, entry1;
-	unsigned char save_control, save_freq_select;
-
-	pin  = find_isa_irq_pin(8, mp_INT);
-	if (pin == -1) {
-		WARN_ON_ONCE(1);
-		return;
-	}
-	apic = find_isa_irq_apic(8, mp_INT);
-	if (apic == -1) {
-		WARN_ON_ONCE(1);
-		return;
-	}
-
-	entry0 = ioapic_read_entry(apic, pin);
-	clear_IO_APIC_pin(apic, pin);
-
-	memset(&entry1, 0, sizeof(entry1));
-
-	entry1.dest_mode = 0;			/* physical delivery */
-	entry1.mask = 0;			/* unmask IRQ now */
-	entry1.dest = hard_smp_processor_id();
-	entry1.delivery_mode = dest_ExtINT;
-	entry1.polarity = entry0.polarity;
-	entry1.trigger = 0;
-	entry1.vector = 0;
-
-	ioapic_write_entry(apic, pin, entry1);
-
-	save_control = CMOS_READ(RTC_CONTROL);
-	save_freq_select = CMOS_READ(RTC_FREQ_SELECT);
-	CMOS_WRITE((save_freq_select & ~RTC_RATE_SELECT) | 0x6,
-		   RTC_FREQ_SELECT);
-	CMOS_WRITE(save_control | RTC_PIE, RTC_CONTROL);
-
-	i = 100;
-	while (i-- > 0) {
-		mdelay(10);
-		if ((CMOS_READ(RTC_INTR_FLAGS) & RTC_PF) == RTC_PF)
-			i -= 10;
-	}
-
-	CMOS_WRITE(save_control, RTC_CONTROL);
-	CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT);
-	clear_IO_APIC_pin(apic, pin);
-
-	ioapic_write_entry(apic, pin, entry0);
-}
-
-static int disable_timer_pin_1 __initdata;
-/* Actually the next is obsolete, but keep it for paranoid reasons -AK */
-static int __init disable_timer_pin_setup(char *arg)
-{
-	disable_timer_pin_1 = 1;
-	return 0;
-}
-early_param("disable_timer_pin_1", disable_timer_pin_setup);
-
-int timer_through_8259 __initdata;
-
-/*
- * This code may look a bit paranoid, but it's supposed to cooperate with
- * a wide range of boards and BIOS bugs.  Fortunately only the timer IRQ
- * is so screwy.  Thanks to Brian Perkins for testing/hacking this beast
- * fanatically on his truly buggy board.
- *
- * FIXME: really need to revamp this for all platforms.
- */
-static inline void __init check_timer(void)
-{
-	struct irq_cfg *cfg = irq_cfg(0);
-	int apic1, pin1, apic2, pin2;
-	unsigned long flags;
-	unsigned int ver;
-	int no_pin1 = 0;
-
-	local_irq_save(flags);
-
-        ver = apic_read(APIC_LVR);
-        ver = GET_APIC_VERSION(ver);
-
-	/*
-	 * get/set the timer IRQ vector:
-	 */
-	disable_8259A_irq(0);
-	assign_irq_vector(0, TARGET_CPUS);
-
-	/*
-	 * As IRQ0 is to be enabled in the 8259A, the virtual
-	 * wire has to be disabled in the local APIC.  Also
-	 * timer interrupts need to be acknowledged manually in
-	 * the 8259A for the i82489DX when using the NMI
-	 * watchdog as that APIC treats NMIs as level-triggered.
-	 * The AEOI mode will finish them in the 8259A
-	 * automatically.
-	 */
-	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
-	init_8259A(1);
-#ifdef CONFIG_X86_32
-	timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver));
-#endif
-
-	pin1  = find_isa_irq_pin(0, mp_INT);
-	apic1 = find_isa_irq_apic(0, mp_INT);
-	pin2  = ioapic_i8259.pin;
-	apic2 = ioapic_i8259.apic;
-
-	apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X "
-		    "apic1=%d pin1=%d apic2=%d pin2=%d\n",
-		    cfg->vector, apic1, pin1, apic2, pin2);
-
-	/*
-	 * Some BIOS writers are clueless and report the ExtINTA
-	 * I/O APIC input from the cascaded 8259A as the timer
-	 * interrupt input.  So just in case, if only one pin
-	 * was found above, try it both directly and through the
-	 * 8259A.
-	 */
-	if (pin1 == -1) {
-#ifdef CONFIG_INTR_REMAP
-		if (intr_remapping_enabled)
-			panic("BIOS bug: timer not connected to IO-APIC");
-#endif
-		pin1 = pin2;
-		apic1 = apic2;
-		no_pin1 = 1;
-	} else if (pin2 == -1) {
-		pin2 = pin1;
-		apic2 = apic1;
-	}
-
-	if (pin1 != -1) {
-		/*
-		 * Ok, does IRQ0 through the IOAPIC work?
-		 */
-		if (no_pin1) {
-			add_pin_to_irq(0, apic1, pin1);
-			setup_timer_IRQ0_pin(apic1, pin1, cfg->vector);
-		}
-		unmask_IO_APIC_irq(0);
-		if (timer_irq_works()) {
-			if (nmi_watchdog == NMI_IO_APIC) {
-				setup_nmi();
-				enable_8259A_irq(0);
-			}
-			if (disable_timer_pin_1 > 0)
-				clear_IO_APIC_pin(0, pin1);
-			goto out;
-		}
-#ifdef CONFIG_INTR_REMAP
-		if (intr_remapping_enabled)
-			panic("timer doesn't work through Interrupt-remapped IO-APIC");
-#endif
-		clear_IO_APIC_pin(apic1, pin1);
-		if (!no_pin1)
-			apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
-				    "8254 timer not connected to IO-APIC\n");
-
-		apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer "
-			    "(IRQ0) through the 8259A ...\n");
-		apic_printk(APIC_QUIET, KERN_INFO
-			    "..... (found apic %d pin %d) ...\n", apic2, pin2);
-		/*
-		 * legacy devices should be connected to IO APIC #0
-		 */
-		replace_pin_at_irq(0, apic1, pin1, apic2, pin2);
-		setup_timer_IRQ0_pin(apic2, pin2, cfg->vector);
-		unmask_IO_APIC_irq(0);
-		enable_8259A_irq(0);
-		if (timer_irq_works()) {
-			apic_printk(APIC_QUIET, KERN_INFO "....... works.\n");
-			timer_through_8259 = 1;
-			if (nmi_watchdog == NMI_IO_APIC) {
-				disable_8259A_irq(0);
-				setup_nmi();
-				enable_8259A_irq(0);
-			}
-			goto out;
-		}
-		/*
-		 * Cleanup, just in case ...
-		 */
-		disable_8259A_irq(0);
-		clear_IO_APIC_pin(apic2, pin2);
-		apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n");
-	}
-
-	if (nmi_watchdog == NMI_IO_APIC) {
-		apic_printk(APIC_QUIET, KERN_WARNING "timer doesn't work "
-			    "through the IO-APIC - disabling NMI Watchdog!\n");
-		nmi_watchdog = NMI_NONE;
-	}
-#ifdef CONFIG_X86_32
-	timer_ack = 0;
-#endif
-
-	apic_printk(APIC_QUIET, KERN_INFO
-		    "...trying to set up timer as Virtual Wire IRQ...\n");
-
-	lapic_register_intr(0);
-	apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector);	/* Fixed mode */
-	enable_8259A_irq(0);
-
-	if (timer_irq_works()) {
-		apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
-		goto out;
-	}
-	disable_8259A_irq(0);
-	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector);
-	apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n");
-
-	apic_printk(APIC_QUIET, KERN_INFO
-		    "...trying to set up timer as ExtINT IRQ...\n");
-
-	init_8259A(0);
-	make_8259A_irq(0);
-	apic_write(APIC_LVT0, APIC_DM_EXTINT);
-
-	unlock_ExtINT_logic();
-
-	if (timer_irq_works()) {
-		apic_printk(APIC_QUIET, KERN_INFO "..... works.\n");
-		goto out;
-	}
-	apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n");
-	panic("IO-APIC + timer doesn't work!  Boot with apic=debug and send a "
-		"report.  Then try booting with the 'noapic' option.\n");
-out:
-	local_irq_restore(flags);
-}
-
-/*
- * Traditionally ISA IRQ2 is the cascade IRQ, and is not available
- * to devices.  However there may be an I/O APIC pin available for
- * this interrupt regardless.  The pin may be left unconnected, but
- * typically it will be reused as an ExtINT cascade interrupt for
- * the master 8259A.  In the MPS case such a pin will normally be
- * reported as an ExtINT interrupt in the MP table.  With ACPI
- * there is no provision for ExtINT interrupts, and in the absence
- * of an override it would be treated as an ordinary ISA I/O APIC
- * interrupt, that is edge-triggered and unmasked by default.  We
- * used to do this, but it caused problems on some systems because
- * of the NMI watchdog and sometimes IRQ0 of the 8254 timer using
- * the same ExtINT cascade interrupt to drive the local APIC of the
- * bootstrap processor.  Therefore we refrain from routing IRQ2 to
- * the I/O APIC in all cases now.  No actual device should request
- * it anyway.  --macro
- */
-#define PIC_IRQS	(1 << PIC_CASCADE_IR)
-
-void __init setup_IO_APIC(void)
-{
-
-#ifdef CONFIG_X86_32
-	enable_IO_APIC();
-#else
-	/*
-	 * calling enable_IO_APIC() is moved to setup_local_APIC for BP
-	 */
-#endif
-
-	io_apic_irqs = ~PIC_IRQS;
-
-	apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
-        /*
-         * Set up IO-APIC IRQ routing.
-         */
-#ifdef CONFIG_X86_32
-        if (!acpi_ioapic)
-                setup_ioapic_ids_from_mpc();
-#endif
-	sync_Arb_IDs();
-	setup_IO_APIC_irqs();
-	init_IO_APIC_traps();
-	check_timer();
-}
-
-/*
- *      Called after all the initialization is done. If we didnt find any
- *      APIC bugs then we can allow the modify fast path
- */
-
-static int __init io_apic_bug_finalize(void)
-{
-        if (sis_apic_bug == -1)
-                sis_apic_bug = 0;
-        return 0;
-}
-
-late_initcall(io_apic_bug_finalize);
-
-struct sysfs_ioapic_data {
-	struct sys_device dev;
-	struct IO_APIC_route_entry entry[0];
-};
-static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
-
-static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
-{
-	struct IO_APIC_route_entry *entry;
-	struct sysfs_ioapic_data *data;
-	int i;
-
-	data = container_of(dev, struct sysfs_ioapic_data, dev);
-	entry = data->entry;
-	for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ )
-		*entry = ioapic_read_entry(dev->id, i);
-
-	return 0;
-}
-
-static int ioapic_resume(struct sys_device *dev)
-{
-	struct IO_APIC_route_entry *entry;
-	struct sysfs_ioapic_data *data;
-	unsigned long flags;
-	union IO_APIC_reg_00 reg_00;
-	int i;
-
-	data = container_of(dev, struct sysfs_ioapic_data, dev);
-	entry = data->entry;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	reg_00.raw = io_apic_read(dev->id, 0);
-	if (reg_00.bits.ID != mp_ioapics[dev->id].mp_apicid) {
-		reg_00.bits.ID = mp_ioapics[dev->id].mp_apicid;
-		io_apic_write(dev->id, 0, reg_00.raw);
-	}
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-	for (i = 0; i < nr_ioapic_registers[dev->id]; i++)
-		ioapic_write_entry(dev->id, i, entry[i]);
-
-	return 0;
-}
-
-static struct sysdev_class ioapic_sysdev_class = {
-	.name = "ioapic",
-	.suspend = ioapic_suspend,
-	.resume = ioapic_resume,
-};
-
-static int __init ioapic_init_sysfs(void)
-{
-	struct sys_device * dev;
-	int i, size, error;
-
-	error = sysdev_class_register(&ioapic_sysdev_class);
-	if (error)
-		return error;
-
-	for (i = 0; i < nr_ioapics; i++ ) {
-		size = sizeof(struct sys_device) + nr_ioapic_registers[i]
-			* sizeof(struct IO_APIC_route_entry);
-		mp_ioapic_data[i] = kzalloc(size, GFP_KERNEL);
-		if (!mp_ioapic_data[i]) {
-			printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
-			continue;
-		}
-		dev = &mp_ioapic_data[i]->dev;
-		dev->id = i;
-		dev->cls = &ioapic_sysdev_class;
-		error = sysdev_register(dev);
-		if (error) {
-			kfree(mp_ioapic_data[i]);
-			mp_ioapic_data[i] = NULL;
-			printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
-			continue;
-		}
-	}
-
-	return 0;
-}
-
-device_initcall(ioapic_init_sysfs);
-
-/*
- * Dynamic irq allocate and deallocation
- */
-unsigned int create_irq(unsigned int irq_want)
-{
-	/* Allocate an unused irq */
-	unsigned int irq;
-	unsigned int new;
-	unsigned long flags;
-	struct irq_cfg *cfg_new;
-
-#ifndef CONFIG_HAVE_SPARSE_IRQ
-	irq_want = nr_irqs - 1;
-#endif
-
-	irq = 0;
-	spin_lock_irqsave(&vector_lock, flags);
-	for (new = irq_want; new > 0; new--) {
-		if (platform_legacy_irq(new))
-			continue;
-		cfg_new = irq_cfg(new);
-		if (cfg_new && cfg_new->vector != 0)
-			continue;
-		/* check if need to create one */
-		if (!cfg_new)
-			cfg_new = irq_cfg_with_new(new);
-		if (__assign_irq_vector(new, TARGET_CPUS) == 0)
-			irq = new;
-		break;
-	}
-	spin_unlock_irqrestore(&vector_lock, flags);
-
-	if (irq > 0) {
-		dynamic_irq_init(irq);
-	}
-	return irq;
-}
-
-void destroy_irq(unsigned int irq)
-{
-	unsigned long flags;
-
-	dynamic_irq_cleanup(irq);
-
-#ifdef CONFIG_INTR_REMAP
-	free_irte(irq);
-#endif
-	spin_lock_irqsave(&vector_lock, flags);
-	__clear_irq_vector(irq);
-	spin_unlock_irqrestore(&vector_lock, flags);
-}
-
-/*
- * MSI message composition
- */
-#ifdef CONFIG_PCI_MSI
-static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
-{
-	struct irq_cfg *cfg;
-	int err;
-	unsigned dest;
-	cpumask_t tmp;
-
-	tmp = TARGET_CPUS;
-	err = assign_irq_vector(irq, tmp);
-	if (err)
-		return err;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, tmp);
-	dest = cpu_mask_to_apicid(tmp);
-
-#ifdef CONFIG_INTR_REMAP
-	if (irq_remapped(irq)) {
-		struct irte irte;
-		int ir_index;
-		u16 sub_handle;
-
-		ir_index = map_irq_to_irte_handle(irq, &sub_handle);
-		BUG_ON(ir_index == -1);
-
-		memset (&irte, 0, sizeof(irte));
-
-		irte.present = 1;
-		irte.dst_mode = INT_DEST_MODE;
-		irte.trigger_mode = 0; /* edge */
-		irte.dlvry_mode = INT_DELIVERY_MODE;
-		irte.vector = cfg->vector;
-		irte.dest_id = IRTE_DEST(dest);
-
-		modify_irte(irq, &irte);
-
-		msg->address_hi = MSI_ADDR_BASE_HI;
-		msg->data = sub_handle;
-		msg->address_lo = MSI_ADDR_BASE_LO | MSI_ADDR_IR_EXT_INT |
-				  MSI_ADDR_IR_SHV |
-				  MSI_ADDR_IR_INDEX1(ir_index) |
-				  MSI_ADDR_IR_INDEX2(ir_index);
-	} else
-#endif
-	{
-		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(cfg->vector);
-	}
-	return err;
-}
-
-#ifdef CONFIG_SMP
-static void set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	struct msi_msg msg;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	read_msi_msg(irq, &msg);
-
-	msg.data &= ~MSI_DATA_VECTOR_MASK;
-	msg.data |= MSI_DATA_VECTOR(cfg->vector);
-	msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
-	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
-
-	write_msi_msg(irq, &msg);
-	desc->affinity = mask;
-}
-
-#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 ir_set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	unsigned int dest;
-	cpumask_t tmp, cleanup_mask;
-	struct irte irte;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (get_irte(irq, &irte))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	irte.vector = cfg->vector;
-	irte.dest_id = IRTE_DEST(dest);
-
-	/*
-	 * atomically update the IRTE with the new destination and vector.
-	 */
-	modify_irte(irq, &irte);
-
-	/*
-	 * After this point, all the interrupts will start arriving
-	 * at the new destination. So, time to cleanup the previous
-	 * vector allocation.
-	 */
-	if (cfg->move_in_progress) {
-		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
-		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
-		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
-		cfg->move_in_progress = 0;
-	}
-
-	desc->affinity = mask;
-}
-#endif
-#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,
-};
-
-#ifdef CONFIG_INTR_REMAP
-static struct irq_chip msi_ir_chip = {
-	.name		= "IR-PCI-MSI",
-	.unmask		= unmask_msi_irq,
-	.mask		= mask_msi_irq,
-	.ack		= ack_x2apic_edge,
-#ifdef CONFIG_SMP
-	.set_affinity	= ir_set_msi_irq_affinity,
-#endif
-	.retrigger	= ioapic_retrigger_irq,
-};
-
-/*
- * Map the PCI dev to the corresponding remapping hardware unit
- * and allocate 'nvec' consecutive interrupt-remapping table entries
- * in it.
- */
-static int msi_alloc_irte(struct pci_dev *dev, int irq, int nvec)
-{
-	struct intel_iommu *iommu;
-	int index;
-
-	iommu = map_dev_to_ir(dev);
-	if (!iommu) {
-		printk(KERN_ERR
-		       "Unable to map PCI %s to iommu\n", pci_name(dev));
-		return -ENOENT;
-	}
-
-	index = alloc_irte(iommu, irq, nvec);
-	if (index < 0) {
-		printk(KERN_ERR
-		       "Unable to allocate %d IRTE for PCI %s\n", nvec,
-		        pci_name(dev));
-		return -ENOSPC;
-	}
-	return index;
-}
-#endif
-
-static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc, int irq)
-{
-	int ret;
-	struct msi_msg msg;
-
-	ret = msi_compose_msg(dev, irq, &msg);
-	if (ret < 0)
-		return ret;
-
-	set_irq_msi(irq, desc);
-	write_msi_msg(irq, &msg);
-
-#ifdef CONFIG_INTR_REMAP
-	if (irq_remapped(irq)) {
-		struct irq_desc *desc = irq_desc(irq);
-		/*
-		 * irq migration in process context
-		 */
-		desc->status |= IRQ_MOVE_PCNTXT;
-		set_irq_chip_and_handler_name(irq, &msi_ir_chip, handle_edge_irq, "edge");
-	} else
-#endif
-		set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
-
-	return 0;
-}
-
-static unsigned int build_irq_for_pci_dev(struct pci_dev *dev)
-{
-	unsigned int irq;
-
-	irq = dev->bus->number;
-	irq <<= 8;
-	irq |= dev->devfn;
-	irq <<= 12;
-
-	return irq;
-}
-
-int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
-{
-	unsigned int irq;
-	int ret;
-	unsigned int irq_want;
-
-	irq_want = build_irq_for_pci_dev(dev) + 0x100;
-
-	irq = create_irq(irq_want);
-	if (irq == 0)
-		return -1;
-
-#ifdef CONFIG_INTR_REMAP
-	if (!intr_remapping_enabled)
-		goto no_ir;
-
-	ret = msi_alloc_irte(dev, irq, 1);
-	if (ret < 0)
-		goto error;
-no_ir:
-#endif
-	ret = setup_msi_irq(dev, desc, irq);
-	if (ret < 0) {
-		destroy_irq(irq);
-		return ret;
-	}
-	return 0;
-
-#ifdef CONFIG_INTR_REMAP
-error:
-	destroy_irq(irq);
-	return ret;
-#endif
-}
-
-int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
-{
-	unsigned int irq;
-	int ret, sub_handle;
-	struct msi_desc *desc;
-	unsigned int irq_want;
-
-#ifdef CONFIG_INTR_REMAP
-	struct intel_iommu *iommu = 0;
-	int index = 0;
-#endif
-
-	irq_want = build_irq_for_pci_dev(dev) + 0x100;
-	sub_handle = 0;
-	list_for_each_entry(desc, &dev->msi_list, list) {
-		irq = create_irq(irq_want--);
-		if (irq == 0)
-			return -1;
-#ifdef CONFIG_INTR_REMAP
-		if (!intr_remapping_enabled)
-			goto no_ir;
-
-		if (!sub_handle) {
-			/*
-			 * allocate the consecutive block of IRTE's
-			 * for 'nvec'
-			 */
-			index = msi_alloc_irte(dev, irq, nvec);
-			if (index < 0) {
-				ret = index;
-				goto error;
-			}
-		} else {
-			iommu = map_dev_to_ir(dev);
-			if (!iommu) {
-				ret = -ENOENT;
-				goto error;
-			}
-			/*
-			 * setup the mapping between the irq and the IRTE
-			 * base index, the sub_handle pointing to the
-			 * appropriate interrupt remap table entry.
-			 */
-			set_irte_irq(irq, iommu, index, sub_handle);
-		}
-no_ir:
-#endif
-		ret = setup_msi_irq(dev, desc, irq);
-		if (ret < 0)
-			goto error;
-		sub_handle++;
-	}
-	return 0;
-
-error:
-	destroy_irq(irq);
-	return ret;
-}
-
-void arch_teardown_msi_irq(unsigned int irq)
-{
-	destroy_irq(irq);
-}
-
-#ifdef CONFIG_DMAR
-#ifdef CONFIG_SMP
-static void dmar_msi_set_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	struct msi_msg msg;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	dmar_msi_read(irq, &msg);
-
-	msg.data &= ~MSI_DATA_VECTOR_MASK;
-	msg.data |= MSI_DATA_VECTOR(cfg->vector);
-	msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
-	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
-
-	dmar_msi_write(irq, &msg);
-	desc->affinity = mask;
-}
-#endif /* CONFIG_SMP */
-
-struct irq_chip dmar_msi_type = {
-	.name = "DMAR_MSI",
-	.unmask = dmar_msi_unmask,
-	.mask = dmar_msi_mask,
-	.ack = ack_apic_edge,
-#ifdef CONFIG_SMP
-	.set_affinity = dmar_msi_set_affinity,
-#endif
-	.retrigger = ioapic_retrigger_irq,
-};
-
-int arch_setup_dmar_msi(unsigned int irq)
-{
-	int ret;
-	struct msi_msg msg;
-
-	ret = msi_compose_msg(NULL, irq, &msg);
-	if (ret < 0)
-		return ret;
-	dmar_msi_write(irq, &msg);
-	set_irq_chip_and_handler_name(irq, &dmar_msi_type, handle_edge_irq,
-		"edge");
-	return 0;
-}
-#endif
-
-#endif /* CONFIG_PCI_MSI */
-/*
- * Hypertransport interrupt support
- */
-#ifdef CONFIG_HT_IRQ
-
-#ifdef CONFIG_SMP
-
-static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector)
-{
-	struct ht_irq_msg msg;
-	fetch_ht_irq_msg(irq, &msg);
-
-	msg.address_lo &= ~(HT_IRQ_LOW_VECTOR_MASK | HT_IRQ_LOW_DEST_ID_MASK);
-	msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK);
-
-	msg.address_lo |= HT_IRQ_LOW_VECTOR(vector) | HT_IRQ_LOW_DEST_ID(dest);
-	msg.address_hi |= HT_IRQ_HIGH_DEST_ID(dest);
-
-	write_ht_irq_msg(irq, &msg);
-}
-
-static void set_ht_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
-{
-	struct irq_cfg *cfg;
-	unsigned int dest;
-	cpumask_t tmp;
-
-	cpus_and(tmp, mask, cpu_online_map);
-	if (cpus_empty(tmp))
-		return;
-
-	if (assign_irq_vector(irq, mask))
-		return;
-
-	cfg = irq_cfg(irq);
-	cpus_and(tmp, cfg->domain, mask);
-	dest = cpu_mask_to_apicid(tmp);
-
-	target_ht_irq(irq, dest, cfg->vector);
-	desc->affinity = 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)
-{
-	struct irq_cfg *cfg;
-	int err;
-	cpumask_t tmp;
-
-	tmp = TARGET_CPUS;
-	err = assign_irq_vector(irq, tmp);
-	if (!err) {
-		struct ht_irq_msg msg;
-		unsigned dest;
-
-		cfg = irq_cfg(irq);
-		cpus_and(tmp, cfg->domain, tmp);
-		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(cfg->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 err;
-}
-#endif /* CONFIG_HT_IRQ */
-
-/* --------------------------------------------------------------------------
-                          ACPI-based IOAPIC Configuration
-   -------------------------------------------------------------------------- */
-
-#ifdef CONFIG_ACPI
-
-#ifdef CONFIG_X86_32
-int __init io_apic_get_unique_id(int ioapic, int apic_id)
-{
-	union IO_APIC_reg_00 reg_00;
-	static physid_mask_t apic_id_map = PHYSID_MASK_NONE;
-	physid_mask_t tmp;
-	unsigned long flags;
-	int i = 0;
-
-	/*
-	 * The P4 platform supports up to 256 APIC IDs on two separate APIC
-	 * buses (one for LAPICs, one for IOAPICs), where predecessors only
-	 * supports up to 16 on one shared APIC bus.
-	 *
-	 * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full
-	 *      advantage of new APIC bus architecture.
-	 */
-
-	if (physids_empty(apic_id_map))
-		apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	reg_00.raw = io_apic_read(ioapic, 0);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-
-	if (apic_id >= get_physical_broadcast()) {
-		printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying "
-			"%d\n", ioapic, apic_id, reg_00.bits.ID);
-		apic_id = reg_00.bits.ID;
-	}
-
-	/*
-	 * Every APIC in a system must have a unique ID or we get lots of nice
-	 * 'stuck on smp_invalidate_needed IPI wait' messages.
-	 */
-	if (check_apicid_used(apic_id_map, apic_id)) {
-
-		for (i = 0; i < get_physical_broadcast(); i++) {
-			if (!check_apicid_used(apic_id_map, i))
-				break;
-		}
-
-		if (i == get_physical_broadcast())
-			panic("Max apic_id exceeded!\n");
-
-		printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
-			"trying %d\n", ioapic, apic_id, i);
-
-		apic_id = i;
-	}
-
-	tmp = apicid_to_cpu_present(apic_id);
-	physids_or(apic_id_map, apic_id_map, tmp);
-
-	if (reg_00.bits.ID != apic_id) {
-		reg_00.bits.ID = apic_id;
-
-		spin_lock_irqsave(&ioapic_lock, flags);
-		io_apic_write(ioapic, 0, reg_00.raw);
-		reg_00.raw = io_apic_read(ioapic, 0);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
-
-		/* Sanity check */
-		if (reg_00.bits.ID != apic_id) {
-			printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
-			return -1;
-		}
-	}
-
-	apic_printk(APIC_VERBOSE, KERN_INFO
-			"IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
-
-	return apic_id;
-}
-
-int __init io_apic_get_version(int ioapic)
-{
-	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);
-
-	return reg_01.bits.version;
-}
-#endif
-
-int __init io_apic_get_redir_entries (int ioapic)
-{
-	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);
-
-	return reg_01.bits.entries;
-}
-
-
-int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
-{
-	if (!IO_APIC_IRQ(irq)) {
-		apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
-			ioapic);
-		return -EINVAL;
-	}
-
-	/*
-	 * IRQs < 16 are already in the irq_2_pin[] map
-	 */
-	if (irq >= 16)
-		add_pin_to_irq(irq, ioapic, pin);
-
-	setup_IO_APIC_irq(ioapic, pin, irq, triggering, polarity);
-
-	return 0;
-}
-
-
-int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
-{
-	int i;
-
-	if (skip_ioapic_setup)
-		return -1;
-
-	for (i = 0; i < mp_irq_entries; i++)
-		if (mp_irqs[i].mp_irqtype == mp_INT &&
-		    mp_irqs[i].mp_srcbusirq == bus_irq)
-			break;
-	if (i >= mp_irq_entries)
-		return -1;
-
-	*trigger = irq_trigger(i);
-	*polarity = irq_polarity(i);
-	return 0;
-}
-
-#endif /* CONFIG_ACPI */
-
-/*
- * This function currently is only a helper for the i386 smp boot process where
- * we need to reprogram the ioredtbls to cater for the cpus which have come online
- * so mask in all cases should simply be TARGET_CPUS
- */
-#ifdef CONFIG_SMP
-void __init setup_ioapic_dest(void)
-{
-	int pin, ioapic, irq, irq_entry;
-	struct irq_cfg *cfg;
-	struct irq_desc *desc;
-
-	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.
-			 */
-			cfg = irq_cfg(irq);
-			if (!cfg->vector)
-				setup_IO_APIC_irq(ioapic, pin, irq,
-						  irq_trigger(irq_entry),
-						  irq_polarity(irq_entry));
-#ifdef CONFIG_INTR_REMAP
-			else if (intr_remapping_enabled) {
-				desc = irq_desc(irq);
-				set_ir_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
-			}
-#endif
-			else {
-				desc = irq_desc(irq);
-				set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
-			}
-		}
-	}
-}
-#endif
-
-#ifdef CONFIG_X86_64
-#define IOAPIC_RESOURCE_NAME_SIZE 11
-
-static struct resource *ioapic_resources;
-
-static struct resource * __init ioapic_setup_resources(void)
-{
-	unsigned long n;
-	struct resource *res;
-	char *mem;
-	int i;
-
-	if (nr_ioapics <= 0)
-		return NULL;
-
-	n = IOAPIC_RESOURCE_NAME_SIZE + sizeof(struct resource);
-	n *= nr_ioapics;
-
-	mem = alloc_bootmem(n);
-	res = (void *)mem;
-
-	if (mem != NULL) {
-		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;
-		}
-	}
-
-	ioapic_resources = res;
-
-	return res;
-}
-#endif
-
-void __init ioapic_init_mappings(void)
-{
-	unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
-	int i;
-#ifdef CONFIG_X86_64
-	struct resource *ioapic_res;
-
-	ioapic_res = ioapic_setup_resources();
-#endif
-	for (i = 0; i < nr_ioapics; i++) {
-		if (smp_found_config) {
-			ioapic_phys = mp_ioapics[i].mp_apicaddr;
-#ifdef CONFIG_X86_32
-                        if (!ioapic_phys) {
-                                printk(KERN_ERR
-                                       "WARNING: bogus zero IO-APIC "
-                                       "address found in MPTABLE, "
-                                       "disabling IO/APIC support!\n");
-                                smp_found_config = 0;
-                                skip_ioapic_setup = 1;
-                                goto fake_ioapic_page;
-                        }
-#endif
-		} else {
-#ifdef CONFIG_X86_32
-fake_ioapic_page:
-#endif
-			ioapic_phys = (unsigned long)
-				alloc_bootmem_pages(PAGE_SIZE);
-			ioapic_phys = __pa(ioapic_phys);
-		}
-		set_fixmap_nocache(idx, ioapic_phys);
-		apic_printk(APIC_VERBOSE,
-			    "mapped IOAPIC to %08lx (%08lx)\n",
-			    __fix_to_virt(idx), ioapic_phys);
-		idx++;
-
-#ifdef CONFIG_X86_64
-		if (ioapic_res != NULL) {
-			ioapic_res->start = ioapic_phys;
-			ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
-			ioapic_res++;
-		}
-#endif
-	}
-}
-
-#ifdef CONFIG_X86_64
-static int __init ioapic_insert_resources(void)
-{
-	int i;
-	struct resource *r = ioapic_resources;
-
-	if (!r) {
-		printk(KERN_ERR
-		       "IO APIC resources could be not be allocated.\n");
-		return -1;
-	}
-
-	for (i = 0; i < nr_ioapics; i++) {
-		insert_resource(&iomem_resource, r);
-		r++;
-	}
-
-	return 0;
-}
-
-/* Insert the IO APIC resources after PCI initialization has occured to handle
- * IO APICS that are mapped in on a BAR in PCI space. */
-late_initcall(ioapic_insert_resources);
-#endif

[PATCH 7/7] make 32 bit have io_apic resource in /proc/iomem

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/x86/kernel/io_apic.c |    8 --------
 1 file changed, 8 deletions(-)

Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic.c
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -3785,7 +3785,6 @@ void __init setup_ioapic_dest(void)
 }
 #endif
 
-#ifdef CONFIG_X86_64
 #define IOAPIC_RESOURCE_NAME_SIZE 11
 
 static struct resource *ioapic_resources;
@@ -3821,17 +3820,14 @@ static struct resource * __init ioapic_s
 
 	return res;
 }
-#endif
 
 void __init ioapic_init_mappings(void)
 {
 	unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
 	int i;
-#ifdef CONFIG_X86_64
 	struct resource *ioapic_res;
 
 	ioapic_res = ioapic_setup_resources();
-#endif
 	for (i = 0; i < nr_ioapics; i++) {
 		if (smp_found_config) {
 			ioapic_phys = mp_ioapics[i].mp_apicaddr;
@@ -3860,17 +3856,14 @@ fake_ioapic_page:
 			    __fix_to_virt(idx), ioapic_phys);
 		idx++;
 
-#ifdef CONFIG_X86_64
 		if (ioapic_res != NULL) {
 			ioapic_res->start = ioapic_phys;
 			ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
 			ioapic_res++;
 		}
-#endif
 	}
 }
 
-#ifdef CONFIG_X86_64
 static int __init ioapic_insert_resources(void)
 {
 	int i;
@@ -3893,4 +3886,3 @@ static int __init ioapic_insert_resource
 /* Insert the IO APIC resources after PCI initialization has occured to handle
  * IO APICS that are mapped in on a BAR in PCI space. */
 late_initcall(ioapic_insert_resources);
-#endif

Re: [PATCH 6/7] rename io_apic_64.c to io_apic.c

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
> 
> ---
>  arch/x86/kernel/Makefile     |    2 
>  arch/x86/kernel/io_apic.c    | 3898 +++++++++++++++++++++++++++++++++++++++++++
>  arch/x86/kernel/io_apic_32.c | 3220 -----------------------------------
>  arch/x86/kernel/io_apic_64.c | 3898 -------------------------------------------
>  4 files changed, 3899 insertions(+), 7119 deletions(-)

very nice stuff!

Could we please make this rename more mechanic? Ie. more intermediate 
changes (or a single change, if it's safe enough) that makes 
io_apic_32.c equal to io_apic_64.c - and _then_ do the rename, ok?

I've looked at the diff (see it below), and there seem to be material 
differences.

Also, the move introduces bogus style that is still present on the 
64-bit side:

-static void mask_IO_APIC_irq(unsigned int irq)
+static void mask_IO_APIC_irq (unsigned int irq)

i'll put it all into tip/irq/sparseirq, to start testing it - but these 
changes _will_ be bisected to so it needs to be as finegrained as 
possible.

	Ingo

1 file changed, 1090 insertions(+), 414 deletions(-)

--- arch/x86/kernel/io_apic_32.c	2008-08-16 09:58:52.000000000 +0200
+++ arch/x86/kernel/io_apic_64.c	2008-08-16 09:58:52.000000000 +0200
@@ -35,7 +35,7 @@
 #include <linux/htirq.h>
 #include <linux/freezer.h>
 #include <linux/kthread.h>
-#include <linux/jiffies.h>      /* time_after() */
+#include <linux/jiffies.h>	/* time_after() */
 #ifdef CONFIG_ACPI
 #include <acpi/acpi_bus.h>
 #endif
@@ -64,8 +64,8 @@
 #define __apicdebuginit(type) static type __init
 
 /*
- *	Is the SiS APIC rmw bug present ?
- *	-1 = don't know, 0 = no, 1 = yes
+ *      Is the SiS APIC rmw bug present ?
+ *      -1 = don't know, 0 = no, 1 = yes
  */
 int sis_apic_bug = -1;
 
@@ -102,7 +102,7 @@ DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BU
 
 int skip_ioapic_setup;
 
-static int __init parse_noapic(char *arg)
+static int __init parse_noapic(char *str)
 {
 	/* disable IO-APIC */
 	disable_ioapic_setup();
@@ -188,7 +188,7 @@ static void __init init_work(void *data)
 	irq_cfgx[legacy_count - 1].next = NULL;
 }
 
-#define for_each_irq_cfg(cfg)           \
+#define for_each_irq_cfg(cfg)		\
 	for (cfg = irq_cfgx; cfg; cfg = cfg->next)
 
 DEFINE_DYN_ARRAY(irq_cfgx, sizeof(struct irq_cfg), nr_irq_cfg, PAGE_SIZE, init_work);
@@ -262,7 +262,6 @@ static struct irq_cfg *irq_cfg_with_new(
 		irq_cfgx = cfg;
 	cfg->irq = irq;
 	printk(KERN_DEBUG "found new irq_cfg for irq %d\n", cfg->irq);
-
 #ifdef CONFIG_HAVE_SPARSE_IRQ_DEBUG
 	{
 		/* dump the results */
@@ -384,9 +383,9 @@ static inline void io_apic_write(unsigne
  */
 static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value)
 {
-	volatile struct io_apic __iomem *io_apic = io_apic_base(apic);
-	if (sis_apic_bug)
-		writel(reg, &io_apic->index);
+	struct io_apic __iomem *io_apic = io_apic_base(apic);
+        if (sis_apic_bug)
+                writel(reg, &io_apic->index);
 	writel(value, &io_apic->data);
 }
 
@@ -494,11 +493,20 @@ static void __target_IO_APIC_irq(unsigne
 
 		apic = entry->apic;
 		pin = entry->pin;
+#ifdef CONFIG_INTR_REMAP
+		/*
+		 * With interrupt-remapping, destination information comes
+		 * from interrupt-remapping table entry.
+		 */
+		if (!irq_remapped(irq))
+			io_apic_write(apic, 0x11 + pin*2, dest);
+#else
 		io_apic_write(apic, 0x11 + pin*2, dest);
+#endif
 		reg = io_apic_read(apic, 0x10 + pin*2);
 		reg &= ~IO_APIC_REDIR_VECTOR_MASK;
 		reg |= vector;
-		io_apic_modify(apic, 0x10 + pin *2, reg);
+		io_apic_modify(apic, 0x10 + pin*2, reg);
 		if (!entry->next)
 			break;
 		entry = entry->next;
@@ -534,7 +542,6 @@ static void set_ioapic_affinity_irq(unsi
 	desc->affinity = mask;
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 }
-
 #endif /* CONFIG_SMP */
 
 /*
@@ -699,7 +706,7 @@ static void __unmask_and_level_IO_APIC_i
 
 #endif
 
-static void mask_IO_APIC_irq(unsigned int irq)
+static void mask_IO_APIC_irq (unsigned int irq)
 {
 	unsigned long flags;
 
@@ -708,7 +715,7 @@ static void mask_IO_APIC_irq(unsigned in
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 }
 
-static void unmask_IO_APIC_irq(unsigned int irq)
+static void unmask_IO_APIC_irq (unsigned int irq)
 {
 	unsigned long flags;
 
@@ -725,14 +732,13 @@ static void clear_IO_APIC_pin(unsigned i
 	entry = ioapic_read_entry(apic, pin);
 	if (entry.delivery_mode == dest_SMI)
 		return;
-
 	/*
 	 * Disable it in the IO-APIC irq-routing table:
 	 */
 	ioapic_mask_entry(apic, pin);
 }
 
-static void clear_IO_APIC(void)
+static void clear_IO_APIC (void)
 {
 	int apic, pin;
 
@@ -741,7 +747,7 @@ static void clear_IO_APIC(void)
 			clear_IO_APIC_pin(apic, pin);
 }
 
-#ifndef CONFIG_SMP
+#if !defined(CONFIG_SMP) && defined(CONFIG_X86_32)
 void send_IPI_self(int vector)
 {
 	unsigned int cfg;
@@ -756,9 +762,9 @@ void send_IPI_self(int vector)
 	 */
 	apic_write(APIC_ICR, cfg);
 }
-#endif /* !CONFIG_SMP */
-
+#endif /* !CONFIG_SMP && CONFIG_X86_32*/
 
+#ifdef CONFIG_X86_32
 /*
  * support for broken MP BIOSs, enables hand-redirection of PIRQ0-7 to
  * specific CPU-side IRQs.
@@ -797,6 +803,75 @@ static int __init ioapic_pirq_setup(char
 }
 
 __setup("pirq=", ioapic_pirq_setup);
+#endif /* CONFIG_X86_32 */
+
+#ifdef CONFIG_INTR_REMAP
+/* I/O APIC RTE contents at the OS boot up */
+static struct IO_APIC_route_entry *early_ioapic_entries[MAX_IO_APICS];
+
+/*
+ * Saves and masks all the unmasked IO-APIC RTE's
+ */
+int save_mask_IO_APIC_setup(void)
+{
+	union IO_APIC_reg_01 reg_01;
+	unsigned long flags;
+	int apic, pin;
+
+	/*
+	 * The number of IO-APIC IRQ registers (== #pins):
+	 */
+	for (apic = 0; apic < nr_ioapics; apic++) {
+		spin_lock_irqsave(&ioapic_lock, flags);
+		reg_01.raw = io_apic_read(apic, 1);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
+	}
+
+	for (apic = 0; apic < nr_ioapics; apic++) {
+		early_ioapic_entries[apic] =
+			kzalloc(sizeof(struct IO_APIC_route_entry) *
+				nr_ioapic_registers[apic], GFP_KERNEL);
+		if (!early_ioapic_entries[apic])
+			return -ENOMEM;
+	}
+
+	for (apic = 0; apic < nr_ioapics; apic++)
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+			struct IO_APIC_route_entry entry;
+
+			entry = early_ioapic_entries[apic][pin] =
+				ioapic_read_entry(apic, pin);
+			if (!entry.mask) {
+				entry.mask = 1;
+				ioapic_write_entry(apic, pin, entry);
+			}
+		}
+	return 0;
+}
+
+void restore_IO_APIC_setup(void)
+{
+	int apic, pin;
+
+	for (apic = 0; apic < nr_ioapics; apic++)
+		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
+			ioapic_write_entry(apic, pin,
+					   early_ioapic_entries[apic][pin]);
+}
+
+void reinit_intr_remapped_IO_APIC(int intr_remapping)
+{
+	/*
+	 * 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();
+}
+#endif
 
 /*
  * Find the IRQ entry number of a certain pin.
@@ -848,7 +923,7 @@ static int __init find_isa_irq_apic(int 
 	}
 	if (i < mp_irq_entries) {
 		int apic;
-		for (apic = 0; apic < nr_ioapics; apic++) {
+		for(apic = 0; apic < nr_ioapics; apic++) {
 			if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic)
 				return apic;
 		}
@@ -867,10 +942,10 @@ int IO_APIC_get_PCI_irq_vector(int bus, 
 {
 	int apic, i, best_guess = -1;
 
-	apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, "
-		"slot:%d, pin:%d.\n", bus, slot, pin);
+	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)) {
-		printk(KERN_WARNING "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
+		apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
 		return -1;
 	}
 	for (i = 0; i < mp_irq_entries; i++) {
@@ -885,7 +960,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, 
 		    !mp_irqs[i].mp_irqtype &&
 		    (bus == lbus) &&
 		    (slot == ((mp_irqs[i].mp_srcbusirq >> 2) & 0x1f))) {
-			int irq = pin_2_irq(i, apic, mp_irqs[i].mp_dstirq);
+			int irq = pin_2_irq(i,apic,mp_irqs[i].mp_dstirq);
 
 			if (!(apic || IO_APIC_IRQ(irq)))
 				continue;
@@ -902,6 +977,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, 
 	}
 	return best_guess;
 }
+
 EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
 
 #if defined(CONFIG_EISA) || defined(CONFIG_MCA)
@@ -918,6 +994,7 @@ static int EISA_ELCR(unsigned int irq)
 			"Broken MPtable reports ISA irq %d\n", irq);
 	return 0;
 }
+
 #endif
 
 /* ISA interrupts are always polarity zero edge triggered,
@@ -931,8 +1008,8 @@ static int EISA_ELCR(unsigned int irq)
  * 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].mp_srcbusirq))
-#define default_EISA_polarity(idx)	default_ISA_polarity(idx)
+#define default_EISA_trigger(idx)       (EISA_ELCR(mp_irqs[idx].mp_srcbusirq))
+#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
 
 /* PCI interrupts are always polarity one level triggered,
  * when listed as conforming in the MP table. */
@@ -940,11 +1017,14 @@ 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. */
+/* 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].mp_srcbusirq))
+#define default_EISA_polarity(idx)      default_ISA_polarity(idx)
 
-#define default_MCA_trigger(idx)	(1)
-#define default_MCA_polarity(idx)	default_ISA_polarity(idx)
 
 static int MPBIOS_polarity(int idx)
 {
@@ -954,36 +1034,36 @@ static int MPBIOS_polarity(int idx)
 	/*
 	 * Determine IRQ line polarity (high active or low active):
 	 */
-	switch (mp_irqs[idx].mp_irqflag & 3) {
-	case 0: /* conforms, ie. bus-type dependent polarity */
-	{
-		polarity = test_bit(bus, mp_bus_not_pci)?
-			default_ISA_polarity(idx):
-			default_PCI_polarity(idx);
-		break;
-	}
-	case 1: /* high active */
+	switch (mp_irqs[idx].mp_irqflag & 3)
 	{
-		polarity = 0;
-		break;
-	}
-	case 2: /* reserved */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		polarity = 1;
-		break;
-	}
-	case 3: /* low active */
-	{
-		polarity = 1;
-		break;
-	}
-	default: /* invalid */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		polarity = 1;
-		break;
-	}
+		case 0: /* conforms, ie. bus-type dependent polarity */
+			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;
+			break;
+		}
+		case 2: /* reserved */
+		{
+			printk(KERN_WARNING "broken BIOS!!\n");
+			polarity = 1;
+			break;
+		}
+		case 3: /* low active */
+		{
+			polarity = 1;
+			break;
+		}
+		default: /* invalid */
+		{
+			printk(KERN_WARNING "broken BIOS!!\n");
+			polarity = 1;
+			break;
+		}
 	}
 	return polarity;
 }
@@ -996,67 +1076,67 @@ static int MPBIOS_trigger(int idx)
 	/*
 	 * Determine IRQ trigger mode (edge or level sensitive):
 	 */
-	switch ((mp_irqs[idx].mp_irqflag>>2) & 3) {
-	case 0: /* conforms, ie. bus-type dependent */
+	switch ((mp_irqs[idx].mp_irqflag>>2) & 3)
 	{
-		trigger = test_bit(bus, mp_bus_not_pci)?
-				default_ISA_trigger(idx):
-				default_PCI_trigger(idx);
+		case 0: /* conforms, ie. bus-type dependent */
+			if (test_bit(bus, mp_bus_not_pci))
+				trigger = default_ISA_trigger(idx);
+			else
+				trigger = default_PCI_trigger(idx);
 #if defined(CONFIG_EISA) || defined(CONFIG_MCA)
-		switch (mp_bus_id_to_type[bus]) {
-		case MP_BUS_ISA: /* ISA pin */
-		{
-			/* set before the switch */
+			switch (mp_bus_id_to_type[bus]) {
+				case MP_BUS_ISA: /* ISA pin */
+				{
+					/* set before the switch */
+					break;
+				}
+				case MP_BUS_EISA: /* EISA pin */
+				{
+					trigger = default_EISA_trigger(idx);
+					break;
+				}
+				case MP_BUS_PCI: /* PCI pin */
+				{
+					/* set before the switch */
+					break;
+				}
+				case MP_BUS_MCA: /* MCA pin */
+				{
+					trigger = default_MCA_trigger(idx);
+					break;
+				}
+				default:
+				{
+					printk(KERN_WARNING "broken BIOS!!\n");
+					trigger = 1;
+					break;
+				}
+			}
+#endif
 			break;
-		}
-		case MP_BUS_EISA: /* EISA pin */
+		case 1: /* edge */
 		{
-			trigger = default_EISA_trigger(idx);
+			trigger = 0;
 			break;
 		}
-		case MP_BUS_PCI: /* PCI pin */
+		case 2: /* reserved */
 		{
-			/* set before the switch */
+			printk(KERN_WARNING "broken BIOS!!\n");
+			trigger = 1;
 			break;
 		}
-		case MP_BUS_MCA: /* MCA pin */
+		case 3: /* level */
 		{
-			trigger = default_MCA_trigger(idx);
+			trigger = 1;
 			break;
 		}
-		default:
+		default: /* invalid */
 		{
 			printk(KERN_WARNING "broken BIOS!!\n");
-			trigger = 1;
+			trigger = 0;
 			break;
 		}
 	}
-#endif
-		break;
-	}
-	case 1: /* edge */
-	{
-		trigger = 0;
-		break;
-	}
-	case 2: /* reserved */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		trigger = 1;
-		break;
-	}
-	case 3: /* level */
-	{
-		trigger = 1;
-		break;
-	}
-	default: /* invalid */
-	{
-		printk(KERN_WARNING "broken BIOS!!\n");
-		trigger = 0;
-		break;
-	}
-	}
 	return trigger;
 }
 
@@ -1082,9 +1162,9 @@ static int pin_2_irq(int idx, int apic, 
 	if (mp_irqs[idx].mp_dstirq != pin)
 		printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n");
 
-	if (test_bit(bus, mp_bus_not_pci))
+	if (test_bit(bus, mp_bus_not_pci)) {
 		irq = mp_irqs[idx].mp_srcbusirq;
-	else {
+	} else {
 		/*
 		 * PCI IRQs are mapped in order
 		 */
@@ -1092,14 +1172,14 @@ static int pin_2_irq(int idx, int apic, 
 		while (i < apic)
 			irq += nr_ioapic_registers[i++];
 		irq += pin;
-
-		/*
-		 * For MPS mode, so far only needed by ES7000 platform
-		 */
-		if (ioapic_renumber_irq)
-			irq = ioapic_renumber_irq(apic, irq);
+                /*
+                 * For MPS mode, so far only needed by ES7000 platform
+                 */
+                if (ioapic_renumber_irq)
+                        irq = ioapic_renumber_irq(apic, irq);
 	}
 
+#ifdef CONFIG_X86_32
 	/*
 	 * PCI IRQ command line redirection. Yes, limits are hardcoded.
 	 */
@@ -1116,6 +1196,8 @@ static int pin_2_irq(int idx, int apic, 
 			}
 		}
 	}
+#endif
+
 	return irq;
 }
 
@@ -1145,74 +1227,70 @@ static int __assign_irq_vector(int irq, 
 	 * Also, we've got to be careful not to trash gate
 	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
 	 */
-        static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0;
-        unsigned int old_vector;
-        int cpu;
-        struct irq_cfg *cfg;
-
-        cfg = irq_cfg(irq);
-
-        /* Only try and allocate irqs on cpus that are present */
-        cpus_and(mask, mask, cpu_online_map);
-
-        if ((cfg->move_in_progress) || cfg->move_cleanup_count)
-                return -EBUSY;
-
-        old_vector = cfg->vector;
-        if (old_vector) {
-                cpumask_t tmp;
-                cpus_and(tmp, cfg->domain, mask);
-                if (!cpus_empty(tmp))
-                        return 0;
-        }
+	static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0;
+	unsigned int old_vector;
+	int cpu;
+	struct irq_cfg *cfg;
+
+	cfg = irq_cfg(irq);
+
+	/* Only try and allocate irqs on cpus that are present */
+	cpus_and(mask, mask, cpu_online_map);
 
-        for_each_cpu_mask_nr(cpu, mask) {
-                cpumask_t domain, new_mask;
-                int new_cpu;
-                int vector, offset;
+	if ((cfg->move_in_progress) || cfg->move_cleanup_count)
+		return -EBUSY;
 
-                domain = vector_allocation_domain(cpu);
-                cpus_and(new_mask, domain, cpu_online_map);
+	old_vector = cfg->vector;
+	if (old_vector) {
+		cpumask_t tmp;
+		cpus_and(tmp, cfg->domain, mask);
+		if (!cpus_empty(tmp))
+			return 0;
+	}
+
+	for_each_cpu_mask_nr(cpu, mask) {
+		cpumask_t domain, new_mask;
+		int new_cpu;
+		int vector, offset;
 
-                vector = current_vector;
-                offset = current_offset;
+		domain = vector_allocation_domain(cpu);
+		cpus_and(new_mask, domain, cpu_online_map);
+
+		vector = current_vector;
+		offset = current_offset;
 next:
-                vector += 8;
-                if (vector >= first_system_vector) {
-                        /* If we run out of vectors on large boxen, must share them. */
-                        offset = (offset + 1) % 8;
-                        vector = FIRST_DEVICE_VECTOR + offset;
-                }
-                if (unlikely(current_vector == vector))
-                        continue;
+		vector += 8;
+		if (vector >= first_system_vector) {
+			/* If we run out of vectors on large boxen, must share them. */
+			offset = (offset + 1) % 8;
+			vector = FIRST_DEVICE_VECTOR + offset;
+		}
+		if (unlikely(current_vector == vector))
+			continue;
 #ifdef CONFIG_X86_64
-                if (vector == IA32_SYSCALL_VECTOR)
-                        goto next;
+		if (vector == IA32_SYSCALL_VECTOR)
+			goto next;
 #else
-                if (vector == SYSCALL_VECTOR)
-                        goto next;
+		if (vector == SYSCALL_VECTOR)
+			goto next;
 #endif
-                for_each_cpu_mask_nr(new_cpu, new_mask)
-                        if (per_cpu(vector_irq, new_cpu)[vector] != -1)
-                                goto next;
-                /* Found one! */
-                current_vector = vector;
-                current_offset = offset;
-                if (old_vector) {
-                        cfg->move_in_progress = 1;
-                        cfg->old_domain = cfg->domain;
-                }
-		printk(KERN_DEBUG "assign_irq_vector: irq %d vector %#x cpu ", irq, vector);
-		for_each_cpu_mask_nr(new_cpu, new_mask) {
-			per_cpu(vector_irq, new_cpu)[vector] = irq;
-			printk(KERN_CONT " %d ", new_cpu);
+		for_each_cpu_mask_nr(new_cpu, new_mask)
+			if (per_cpu(vector_irq, new_cpu)[vector] != -1)
+				goto next;
+		/* Found one! */
+		current_vector = vector;
+		current_offset = offset;
+		if (old_vector) {
+			cfg->move_in_progress = 1;
+			cfg->old_domain = cfg->domain;
 		}
-		printk(KERN_CONT "\n");
-                cfg->vector = vector;
-                cfg->domain = domain;
-                return 0;
-        }
-        return -ENOSPC;
+		for_each_cpu_mask_nr(new_cpu, new_mask)
+			per_cpu(vector_irq, new_cpu)[vector] = irq;
+		cfg->vector = vector;
+		cfg->domain = domain;
+		return 0;
+	}
+	return -ENOSPC;
 }
 
 static int assign_irq_vector(int irq, cpumask_t mask)
@@ -1223,7 +1301,6 @@ static int assign_irq_vector(int irq, cp
 	spin_lock_irqsave(&vector_lock, flags);
 	err = __assign_irq_vector(irq, mask);
 	spin_unlock_irqrestore(&vector_lock, flags);
-
 	return err;
 }
 
@@ -1269,36 +1346,39 @@ void __setup_vector_irq(int cpu)
 		cfg = irq_cfg(irq);
 		if (!cpu_isset(cpu, cfg->domain))
 			per_cpu(vector_irq, cpu)[vector] = -1;
-        }
+	}
 }
 
 static struct irq_chip ioapic_chip;
+#ifdef CONFIG_INTR_REMAP
+static struct irq_chip ir_ioapic_chip;
+#endif
 
-#define IOAPIC_AUTO	-1
-#define IOAPIC_EDGE	0
-#define IOAPIC_LEVEL	1
+#define IOAPIC_AUTO     -1
+#define IOAPIC_EDGE     0
+#define IOAPIC_LEVEL    1
 
 #ifdef CONFIG_X86_32
 static inline int IO_APIC_irq_trigger(int irq)
 {
-	int apic, idx, pin;
+        int apic, idx, pin;
 
-	for (apic = 0; apic < nr_ioapics; apic++) {
-		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
-			idx = find_irq_entry(apic, pin, mp_INT);
-			if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
-				return irq_trigger(idx);
-		}
-	}
-	/*
-	 * nonexistent IRQs are edge default
-	 */
-	return 0;
+        for (apic = 0; apic < nr_ioapics; apic++) {
+                for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
+                        idx = find_irq_entry(apic, pin, mp_INT);
+                        if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin)))
+                                return irq_trigger(idx);
+                }
+        }
+        /*
+         * nonexistent IRQs are edge default
+         */
+        return 0;
 }
 #else
 static inline int IO_APIC_irq_trigger(int irq)
 {
-        return 1;
+	return 1;
 }
 #endif
 
@@ -1318,13 +1398,27 @@ static void ioapic_register_intr(int irq
 	else
 		desc->status &= ~IRQ_LEVEL;
 
+#ifdef CONFIG_INTR_REMAP
+	if (irq_remapped(irq)) {
+		desc->status |= IRQ_MOVE_PCNTXT;
+		if (trigger)
+			set_irq_chip_and_handler_name(irq, &ir_ioapic_chip,
+						      handle_fasteoi_irq,
+						     "fasteoi");
+		else
+			set_irq_chip_and_handler_name(irq, &ir_ioapic_chip,
+						      handle_edge_irq, "edge");
+		return;
+	}
+#endif
 	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
 	    trigger == IOAPIC_LEVEL)
 		set_irq_chip_and_handler_name(irq, &ioapic_chip,
-					 handle_fasteoi_irq, "fasteoi");
+					      handle_fasteoi_irq,
+					      "fasteoi");
 	else
 		set_irq_chip_and_handler_name(irq, &ioapic_chip,
-					 handle_edge_irq, "edge");
+					      handle_edge_irq, "edge");
 }
 
 static int setup_ioapic_entry(int apic, int irq,
@@ -1337,11 +1431,45 @@ static int setup_ioapic_entry(int apic, 
 	 */
 	memset(entry,0,sizeof(*entry));
 
-	entry->delivery_mode = INT_DELIVERY_MODE;
-	entry->dest_mode = INT_DEST_MODE;
-	entry->dest = destination;
+#ifdef CONFIG_INTR_REMAP
+	if (intr_remapping_enabled) {
+		struct intel_iommu *iommu = map_ioapic_to_ir(apic);
+		struct irte irte;
+		struct IR_IO_APIC_route_entry *ir_entry =
+			(struct IR_IO_APIC_route_entry *) entry;
+		int index;
+
+		if (!iommu)
+			panic("No mapping iommu for ioapic %d\n", apic);
+
+		index = alloc_irte(iommu, irq, 1);
+		if (index < 0)
+			panic("Failed to allocate IRTE for ioapic %d\n", apic);
+
+		memset(&irte, 0, sizeof(irte));
+
+		irte.present = 1;
+		irte.dst_mode = INT_DEST_MODE;
+		irte.trigger_mode = trigger;
+		irte.dlvry_mode = INT_DELIVERY_MODE;
+		irte.vector = vector;
+		irte.dest_id = IRTE_DEST(destination);
+
+		modify_irte(irq, &irte);
+
+		ir_entry->index2 = (index >> 15) & 0x1;
+		ir_entry->zero = 0;
+		ir_entry->format = 1;
+		ir_entry->index = (index & 0x7fff);
+	} else
+#endif
+	{
+		entry->delivery_mode = INT_DELIVERY_MODE;
+		entry->dest_mode = INT_DEST_MODE;
+		entry->dest = destination;
+	}
 
-	entry->mask = 0;                                /* enable IRQ */
+	entry->mask = 0;				/* enable IRQ */
 	entry->trigger = trigger;
 	entry->polarity = polarity;
 	entry->vector = vector;
@@ -1351,12 +1479,11 @@ static int setup_ioapic_entry(int apic, 
 	 */
 	if (trigger)
 		entry->mask = 1;
-
 	return 0;
 }
 
 static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq,
-                              int trigger, int polarity)
+			      int trigger, int polarity)
 {
 	struct irq_cfg *cfg;
 	struct IO_APIC_route_entry entry;
@@ -1420,10 +1547,10 @@ static void __init setup_IO_APIC_irqs(vo
 		}
 
 		irq = pin_2_irq(idx, apic, pin);
-
+#ifdef CONFIG_X86_32
                 if (multi_timer_check(apic, irq))
                         continue;
-
+#endif
 		add_pin_to_irq(irq, apic, pin);
 
 		setup_IO_APIC_irq(apic, pin, irq,
@@ -1443,6 +1570,11 @@ static void __init setup_timer_IRQ0_pin(
 {
 	struct IO_APIC_route_entry entry;
 
+#ifdef CONFIG_INTR_REMAP
+	if (intr_remapping_enabled)
+		return;
+#endif
+
 	memset(&entry, 0, sizeof(entry));
 
 	/*
@@ -1461,7 +1593,7 @@ static void __init setup_timer_IRQ0_pin(
 	 * The timer IRQ doesn't have to know that behind the
 	 * scene we may have a 8259A-master in AEOI mode ...
 	 */
-	ioapic_register_intr(0, IOAPIC_EDGE);
+	set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge");
 
 	/*
 	 * Add it to the IO-APIC irq-routing table:
@@ -1501,17 +1633,18 @@ __apicdebuginit(void) print_IO_APIC(void
 	reg_01.raw = io_apic_read(apic, 1);
 	if (reg_01.bits.version >= 0x10)
 		reg_02.raw = io_apic_read(apic, 2);
-	if (reg_01.bits.version >= 0x20)
-		reg_03.raw = io_apic_read(apic, 3);
+        if (reg_01.bits.version >= 0x20)
+                reg_03.raw = io_apic_read(apic, 3);
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 
+	printk("\n");
 	printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid);
 	printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
 	printk(KERN_DEBUG ".......    : physical APIC id: %02X\n", reg_00.bits.ID);
 	printk(KERN_DEBUG ".......    : Delivery Type: %X\n", reg_00.bits.delivery_type);
 	printk(KERN_DEBUG ".......    : LTS          : %X\n", reg_00.bits.LTS);
 
-	printk(KERN_DEBUG ".... register #01: %08X\n", reg_01.raw);
+	printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)&reg_01);
 	printk(KERN_DEBUG ".......     : max redirection entries: %04X\n", reg_01.bits.entries);
 
 	printk(KERN_DEBUG ".......     : PRQ implemented: %X\n", reg_01.bits.PRQ);
@@ -1548,7 +1681,10 @@ __apicdebuginit(void) print_IO_APIC(void
 
 		entry = ioapic_read_entry(apic, i);
 
-		printk(KERN_DEBUG " %02x %02X  ", i, entry.dest);
+		printk(KERN_DEBUG " %02x %03X ",
+			i,
+			entry.dest
+		);
 
 		printk("%1d    %1d    %1d   %1d   %1d    %1d    %1d    %02X\n",
 			entry.mask,
@@ -1567,7 +1703,7 @@ __apicdebuginit(void) print_IO_APIC(void
 		struct irq_pin_list *entry = cfg->irq_2_pin;
 		if (!entry)
 			continue;
-		printk(KERN_DEBUG "IRQ%d ", i);
+		printk(KERN_DEBUG "IRQ%d ", cfg->irq);
 		for (;;) {
 			printk("-> %d:%d", entry->apic, entry->pin);
 			if (!entry->next)
@@ -1614,8 +1750,7 @@ __apicdebuginit(void) print_local_APIC(v
 	printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
 		smp_processor_id(), hard_smp_processor_id());
 	v = apic_read(APIC_ID);
-	printk(KERN_INFO "... APIC ID:      %08x (%01x)\n", v,
-			GET_APIC_ID(v));
+	printk(KERN_INFO "... APIC ID:      %08x (%01x)\n", v, read_apic_id());
 	v = apic_read(APIC_LVR);
 	printk(KERN_INFO "... APIC VERSION: %08x\n", v);
 	ver = GET_APIC_VERSION(v);
@@ -1624,7 +1759,7 @@ __apicdebuginit(void) print_local_APIC(v
 	v = apic_read(APIC_TASKPRI);
 	printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
 
-	if (APIC_INTEGRATED(ver)) {			/* !82489DX */
+	if (APIC_INTEGRATED(ver)) {                     /* !82489DX */
 		v = apic_read(APIC_ARBPRI);
 		printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
 			v & APIC_ARBPRI_MASK);
@@ -1650,9 +1785,10 @@ __apicdebuginit(void) print_local_APIC(v
 	printk(KERN_DEBUG "... APIC IRR field:\n");
 	print_APIC_bitfield(APIC_IRR);
 
-	if (APIC_INTEGRATED(ver)) {		/* !82489DX */
-		if (maxlvt > 3)		/* Due to the Pentium erratum 3AP. */
+	if (APIC_INTEGRATED(ver)) {             /* !82489DX */
+		if (maxlvt > 3)         /* Due to the Pentium erratum 3AP. */
 			apic_write(APIC_ESR, 0);
+
 		v = apic_read(APIC_ESR);
 		printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
 	}
@@ -1710,11 +1846,11 @@ __apicdebuginit(void) print_PIC(void)
 	v = inb(0xa0) << 8 | inb(0x20);
 	printk(KERN_DEBUG "... PIC  IRR: %04x\n", v);
 
-	outb(0x0b, 0xa0);
-	outb(0x0b, 0x20);
+	outb(0x0b,0xa0);
+	outb(0x0b,0x20);
 	v = inb(0xa0) << 8 | inb(0x20);
-	outb(0x0a, 0xa0);
-	outb(0x0a, 0x20);
+	outb(0x0a,0xa0);
+	outb(0x0a,0x20);
 
 	spin_unlock_irqrestore(&i8259A_lock, flags);
 
@@ -1739,16 +1875,19 @@ fs_initcall(print_all_ICs);
 /* Where if anywhere is the i8259 connect in external int mode */
 static struct { int pin, apic; } ioapic_i8259 = { -1, -1 };
 
-static void __init enable_IO_APIC(void)
+void __init enable_IO_APIC(void)
 {
 	union IO_APIC_reg_01 reg_01;
 	int i8259_apic, i8259_pin;
-	int i, apic;
+	int apic;
 	unsigned long flags;
 
+#ifdef CONFIG_X86_32
+	int i;
 	if (!pirqs_enabled)
 		for (i = 0; i < MAX_PIRQS; i++)
 			pirq_entries[i] = -1;
+#endif
 
 	/*
 	 * The number of IO-APIC IRQ registers (== #pins):
@@ -1759,7 +1898,7 @@ static void __init enable_IO_APIC(void)
 		spin_unlock_irqrestore(&ioapic_lock, flags);
 		nr_ioapic_registers[apic] = reg_01.bits.entries+1;
 	}
-	for (apic = 0; apic < nr_ioapics; apic++) {
+	for(apic = 0; apic < nr_ioapics; apic++) {
 		int pin;
 		/* See if any of the pins is in ExtINT mode */
 		for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
@@ -1830,16 +1969,18 @@ void disable_IO_APIC(void)
 		entry.dest_mode       = 0; /* Physical */
 		entry.delivery_mode   = dest_ExtINT; /* ExtInt */
 		entry.vector          = 0;
-		entry.dest	      = read_apic_id();
+		entry.dest            = read_apic_id();
 
 		/*
 		 * Add it to the IO-APIC irq-routing table:
 		 */
 		ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry);
 	}
+
 	disconnect_bsp_APIC(ioapic_i8259.pin != -1);
 }
 
+#ifdef CONFIG_X86_32
 /*
  * function to set the IO-APIC physical IDs based on the
  * values stored in the MPC table.
@@ -1940,8 +2081,6 @@ static void __init setup_ioapic_ids_from
 
 		reg_00.bits.ID = mp_ioapics[apic].mp_apicid;
 		spin_lock_irqsave(&ioapic_lock, flags);
-		io_apic_write(apic, 0, reg_00.raw);
-		spin_unlock_irqrestore(&ioapic_lock, flags);
 
 		/*
 		 * Sanity check
@@ -1955,6 +2094,7 @@ static void __init setup_ioapic_ids_from
 			apic_printk(APIC_VERBOSE, " ok.\n");
 	}
 }
+#endif
 
 int no_timer_check __initdata;
 
@@ -1994,9 +2134,10 @@ static int __init timer_irq_works(void)
 	 * might have cached one ExtINT interrupt.  Finally, at
 	 * least one tick may be lost due to delays.
 	 */
+
+	/* jiffies wrap? */
 	if (time_after(jiffies, t1 + 4))
 		return 1;
-
 	return 0;
 }
 
@@ -2014,8 +2155,6 @@ static int __init timer_irq_works(void)
  */
 
 /*
- * Startup quirk:
- *
  * Starting up a edge-triggered IO-APIC interrupt is
  * nasty - we need to make sure that we get the edge.
  * If it is already asserted for some reason, we need
@@ -2023,9 +2162,8 @@ static int __init timer_irq_works(void)
  *
  * This is not complete - we should be able to fake
  * an edge even if it isn't on the 8259A...
- *
- * (We do this for level-triggered IRQs too - it cannot hurt.)
  */
+
 static unsigned int startup_ioapic_irq(unsigned int irq)
 {
 	int was_pending = 0;
@@ -2043,58 +2181,229 @@ static unsigned int startup_ioapic_irq(u
 	return was_pending;
 }
 
+#ifdef CONFIG_X86_64
 static int ioapic_retrigger_irq(unsigned int irq)
 {
-	send_IPI_self(irq_cfg(irq)->vector);
+
+	struct irq_cfg *cfg = irq_cfg(irq);
+	unsigned long flags;
+
+	spin_lock_irqsave(&vector_lock, flags);
+	send_IPI_mask(cpumask_of_cpu(first_cpu(cfg->domain)), cfg->vector);
+	spin_unlock_irqrestore(&vector_lock, flags);
 
 	return 1;
 }
-
-#ifdef CONFIG_SMP
-asmlinkage void smp_irq_move_cleanup_interrupt(void)
+#else
+static int ioapic_retrigger_irq(unsigned int irq)
 {
-	unsigned vector, me;
-	ack_APIC_irq();
-	irq_enter();
-
-	me = smp_processor_id();
-	for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
-		unsigned int irq;
-		struct irq_desc *desc;
-		struct irq_cfg *cfg;
-		irq = __get_cpu_var(vector_irq)[vector];
+        send_IPI_self(irq_cfg(irq)->vector);
 
-		desc = irq_desc(irq);
-		if (!desc)
-			continue;
+        return 1;
+}
+#endif
 
-		cfg = irq_cfg(irq);
-		spin_lock(&desc->lock);
-		if (!cfg->move_cleanup_count)
-			goto unlock;
+/*
+ * 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.
+ */
 
-		if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
-			goto unlock;
+#ifdef CONFIG_SMP
 
-		__get_cpu_var(vector_irq)[vector] = -1;
-		cfg->move_cleanup_count--;
-unlock:
-		spin_unlock(&desc->lock);
-	}
+#ifdef CONFIG_INTR_REMAP
+static void ir_irq_migration(struct work_struct *work);
 
-	irq_exit();
-}
+static DECLARE_DELAYED_WORK(ir_migration_work, ir_irq_migration);
 
-static void irq_complete_move(unsigned int irq)
+/*
+ * Migrate the IO-APIC irq in the presence of intr-remapping.
+ *
+ * For edge triggered, irq migration is a simple atomic update(of vector
+ * and cpu destination) of IRTE and flush the hardware cache.
+ *
+ * For level triggered, we need to modify the io-apic RTE aswell with the update
+ * vector information, along with modifying IRTE with vector and destination.
+ * So irq migration for level triggered is little  bit more complex compared to
+ * edge triggered migration. But the good news is, we use the same algorithm
+ * for level triggered migration as we have today, only difference being,
+ * we now initiate the irq migration from process context instead of the
+ * interrupt context.
+ *
+ * In future, when we do a directed EOI (combined with cpu EOI broadcast
+ * suppression) to the IO-APIC, level triggered irq migration will also be
+ * as simple as edge triggered migration and we can do the irq migration
+ * with a simple atomic update to IO-APIC RTE.
+ */
+static void migrate_ioapic_irq(int irq, cpumask_t mask)
 {
-	struct irq_cfg *cfg = irq_cfg(irq);
-	unsigned vector, me;
+	struct irq_cfg *cfg;
+	struct irq_desc *desc;
+	cpumask_t tmp, cleanup_mask;
+	struct irte irte;
+	int modify_ioapic_rte;
+	unsigned int dest;
+	unsigned long flags;
 
-	if (likely(!cfg->move_in_progress))
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
 		return;
 
-	vector = ~get_irq_regs()->orig_ax;
-	me = smp_processor_id();
+	if (get_irte(irq, &irte))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	desc = irq_desc(irq);
+	modify_ioapic_rte = desc->status & IRQ_LEVEL;
+	if (modify_ioapic_rte) {
+		spin_lock_irqsave(&ioapic_lock, flags);
+		__target_IO_APIC_irq(irq, dest, cfg->vector);
+		spin_unlock_irqrestore(&ioapic_lock, flags);
+	}
+
+	irte.vector = cfg->vector;
+	irte.dest_id = IRTE_DEST(dest);
+
+	/*
+	 * Modified the IRTE and flushes the Interrupt entry cache.
+	 */
+	modify_irte(irq, &irte);
+
+	if (cfg->move_in_progress) {
+		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
+		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
+		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+		cfg->move_in_progress = 0;
+	}
+
+	desc->affinity = mask;
+}
+
+static int migrate_irq_remapped_level(int irq)
+{
+	int ret = -1;
+	struct irq_desc *desc = irq_desc(irq);
+
+	mask_IO_APIC_irq(irq);
+
+	if (io_apic_level_ack_pending(irq)) {
+		/*
+	 	 * Interrupt in progress. Migrating irq now will change the
+		 * vector information in the IO-APIC RTE and that will confuse
+		 * the EOI broadcast performed by cpu.
+		 * So, delay the irq migration to the next instance.
+		 */
+		schedule_delayed_work(&ir_migration_work, 1);
+		goto unmask;
+	}
+
+	/* everthing is clear. we have right of way */
+	migrate_ioapic_irq(irq, desc->pending_mask);
+
+	ret = 0;
+	desc->status &= ~IRQ_MOVE_PENDING;
+	cpus_clear(desc->pending_mask);
+
+unmask:
+	unmask_IO_APIC_irq(irq);
+	return ret;
+}
+
+static void ir_irq_migration(struct work_struct *work)
+{
+	unsigned int irq;
+	struct irq_desc *desc;
+
+	for_each_irq_desc(irq, desc) {
+		if (desc->status & IRQ_MOVE_PENDING) {
+			unsigned long flags;
+
+			spin_lock_irqsave(&desc->lock, flags);
+			if (!desc->chip->set_affinity ||
+			    !(desc->status & IRQ_MOVE_PENDING)) {
+				desc->status &= ~IRQ_MOVE_PENDING;
+				spin_unlock_irqrestore(&desc->lock, flags);
+				continue;
+			}
+
+			desc->chip->set_affinity(irq, desc, desc->pending_mask);
+			spin_unlock_irqrestore(&desc->lock, flags);
+		}
+	}
+}
+
+/*
+ * Migrates the IRQ destination in the process context.
+ */
+static void set_ir_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	if (desc->status & IRQ_LEVEL) {
+		desc->status |= IRQ_MOVE_PENDING;
+		desc->pending_mask = mask;
+		migrate_irq_remapped_level(irq);
+		return;
+	}
+
+	migrate_ioapic_irq(irq, mask);
+}
+#endif
+
+asmlinkage void smp_irq_move_cleanup_interrupt(void)
+{
+	unsigned vector, me;
+	ack_APIC_irq();
+#ifdef CONFIG_X86_64
+	exit_idle();
+#endif
+	irq_enter();
+
+	me = smp_processor_id();
+	for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
+		unsigned int irq;
+		struct irq_desc *desc;
+		struct irq_cfg *cfg;
+		irq = __get_cpu_var(vector_irq)[vector];
+
+		desc = irq_desc(irq);
+		if (!desc)
+			continue;
+
+		cfg = irq_cfg(irq);
+		spin_lock(&desc->lock);
+		if (!cfg->move_cleanup_count)
+			goto unlock;
+
+		if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
+			goto unlock;
+
+		__get_cpu_var(vector_irq)[vector] = -1;
+		cfg->move_cleanup_count--;
+unlock:
+		spin_unlock(&desc->lock);
+	}
+
+	irq_exit();
+}
+
+static void irq_complete_move(unsigned int irq)
+{
+	struct irq_cfg *cfg = irq_cfg(irq);
+	unsigned vector, me;
+
+	if (likely(!cfg->move_in_progress))
+		return;
+
+	vector = ~get_irq_regs()->orig_ax;
+	me = smp_processor_id();
 	if ((vector == cfg->vector) && cpu_isset(me, cfg->domain)) {
 		cpumask_t cleanup_mask;
 
@@ -2107,6 +2416,17 @@ static void irq_complete_move(unsigned i
 #else
 static inline void irq_complete_move(unsigned int irq) {}
 #endif
+#ifdef CONFIG_INTR_REMAP
+static void ack_x2apic_level(unsigned int irq, struct irq_desc *desc)
+{
+	ack_x2APIC_irq();
+}
+
+static void ack_x2apic_edge(unsigned int irq, struct irq_desc *desc)
+{
+	ack_x2APIC_irq();
+}
+#endif
 
 static void ack_apic_edge(unsigned int irq, struct irq_desc *desc)
 {
@@ -2118,55 +2438,55 @@ static void ack_apic_edge(unsigned int i
 #ifdef CONFIG_X86_64
 static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
 {
-        int do_unmask_irq = 0;
+	int do_unmask_irq = 0;
 
-        irq_complete_move(irq);
+	irq_complete_move(irq);
 #ifdef CONFIG_GENERIC_PENDING_IRQ
-        /* If we are moving the irq we need to mask it */
-        if (unlikely(desc->status & IRQ_MOVE_PENDING)) {
-                do_unmask_irq = 1;
-                mask_IO_APIC_irq(irq);
-        }
+	/* If we are moving the irq we need to mask it */
+	if (unlikely(desc->status & IRQ_MOVE_PENDING)) {
+		do_unmask_irq = 1;
+		mask_IO_APIC_irq(irq);
+	}
 #endif
 
-        /*
-         * We must acknowledge the irq before we move it or the acknowledge will
-         * not propagate properly.
-         */
-        ack_APIC_irq();
+	/*
+	 * We must acknowledge the irq before we move it or the acknowledge will
+	 * not propagate properly.
+	 */
+	ack_APIC_irq();
 
-        /* Now we can move and renable the irq */
-        if (unlikely(do_unmask_irq)) {
-                /* Only migrate the irq if the ack has been received.
-                 *
-                 * On rare occasions the broadcast level triggered ack gets
-                 * delayed going to ioapics, and if we reprogram the
-                 * vector while Remote IRR is still set the irq will never
-                 * fire again.
-                 *
-                 * To prevent this scenario we read the Remote IRR bit
-                 * of the ioapic.  This has two effects.
-                 * - On any sane system the read of the ioapic will
-                 *   flush writes (and acks) going to the ioapic from
-                 *   this cpu.
-                 * - We get to see if the ACK has actually been delivered.
-                 *
-                 * Based on failed experiments of reprogramming the
-                 * ioapic entry from outside of irq context starting
-                 * with masking the ioapic entry and then polling until
-                 * Remote IRR was clear before reprogramming the
-                 * ioapic I don't trust the Remote IRR bit to be
-                 * completey accurate.
-                 *
-                 * However there appears to be no other way to plug
-                 * this race, so if the Remote IRR bit is not
-                 * accurate and is causing problems then it is a hardware bug
-                 * and you can go talk to the chipset vendor about it.
-                 */
-                if (!io_apic_level_ack_pending(irq))
-                        move_masked_irq(irq, desc);
-                unmask_IO_APIC_irq(irq);
-        }
+	/* Now we can move and renable the irq */
+	if (unlikely(do_unmask_irq)) {
+		/* Only migrate the irq if the ack has been received.
+		 *
+		 * On rare occasions the broadcast level triggered ack gets
+		 * delayed going to ioapics, and if we reprogram the
+		 * vector while Remote IRR is still set the irq will never
+		 * fire again.
+		 *
+		 * To prevent this scenario we read the Remote IRR bit
+		 * of the ioapic.  This has two effects.
+		 * - On any sane system the read of the ioapic will
+		 *   flush writes (and acks) going to the ioapic from
+		 *   this cpu.
+		 * - We get to see if the ACK has actually been delivered.
+		 *
+		 * Based on failed experiments of reprogramming the
+		 * ioapic entry from outside of irq context starting
+		 * with masking the ioapic entry and then polling until
+		 * Remote IRR was clear before reprogramming the
+		 * ioapic I don't trust the Remote IRR bit to be
+		 * completey accurate.
+		 *
+		 * However there appears to be no other way to plug
+		 * this race, so if the Remote IRR bit is not
+		 * accurate and is causing problems then it is a hardware bug
+		 * and you can go talk to the chipset vendor about it.
+		 */
+		if (!io_apic_level_ack_pending(irq))
+			move_masked_irq(irq, desc);
+		unmask_IO_APIC_irq(irq);
+	}
 }
 #else
 atomic_t irq_mis_count;
@@ -2177,25 +2497,25 @@ static void ack_apic_level(unsigned int 
 
 	irq_complete_move(irq);
 	move_native_irq(irq, desc);
-/*
- * 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
+	*/
 	i = irq_cfg(irq)->vector;
 
 	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
@@ -2225,6 +2545,20 @@ static struct irq_chip ioapic_chip __rea
 	.retrigger	= ioapic_retrigger_irq,
 };
 
+#ifdef CONFIG_INTR_REMAP
+static struct irq_chip ir_ioapic_chip __read_mostly = {
+	.name 		= "IR-IO-APIC",
+	.startup 	= startup_ioapic_irq,
+	.mask	 	= mask_IO_APIC_irq,
+	.unmask	 	= unmask_IO_APIC_irq,
+	.ack 		= ack_x2apic_edge,
+	.eoi 		= ack_x2apic_level,
+#ifdef CONFIG_SMP
+	.set_affinity 	= set_ir_ioapic_affinity_irq,
+#endif
+	.retrigger	= ioapic_retrigger_irq,
+};
+#endif
 
 static inline void init_IO_APIC_traps(void)
 {
@@ -2383,12 +2717,12 @@ static inline void __init unlock_ExtINT_
 
 static int disable_timer_pin_1 __initdata;
 /* Actually the next is obsolete, but keep it for paranoid reasons -AK */
-static int __init parse_disable_timer_pin_1(char *arg)
+static int __init disable_timer_pin_setup(char *arg)
 {
 	disable_timer_pin_1 = 1;
 	return 0;
 }
-early_param("disable_timer_pin_1", parse_disable_timer_pin_1);
+early_param("disable_timer_pin_1", disable_timer_pin_setup);
 
 int timer_through_8259 __initdata;
 
@@ -2397,6 +2731,8 @@ int timer_through_8259 __initdata;
  * a wide range of boards and BIOS bugs.  Fortunately only the timer IRQ
  * is so screwy.  Thanks to Brian Perkins for testing/hacking this beast
  * fanatically on his truly buggy board.
+ *
+ * FIXME: really need to revamp this for all platforms.
  */
 static inline void __init check_timer(void)
 {
@@ -2408,8 +2744,8 @@ static inline void __init check_timer(vo
 
 	local_irq_save(flags);
 
-	ver = apic_read(APIC_LVR);
-	ver = GET_APIC_VERSION(ver);
+        ver = apic_read(APIC_LVR);
+        ver = GET_APIC_VERSION(ver);
 
 	/*
 	 * get/set the timer IRQ vector:
@@ -2428,7 +2764,9 @@ static inline void __init check_timer(vo
 	 */
 	apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
 	init_8259A(1);
+#ifdef CONFIG_X86_32
 	timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver));
+#endif
 
 	pin1  = find_isa_irq_pin(0, mp_INT);
 	apic1 = find_isa_irq_apic(0, mp_INT);
@@ -2447,6 +2785,10 @@ static inline void __init check_timer(vo
 	 * 8259A.
 	 */
 	if (pin1 == -1) {
+#ifdef CONFIG_INTR_REMAP
+		if (intr_remapping_enabled)
+			panic("BIOS bug: timer not connected to IO-APIC");
+#endif
 		pin1 = pin2;
 		apic1 = apic2;
 		no_pin1 = 1;
@@ -2473,6 +2815,10 @@ static inline void __init check_timer(vo
 				clear_IO_APIC_pin(0, pin1);
 			goto out;
 		}
+#ifdef CONFIG_INTR_REMAP
+		if (intr_remapping_enabled)
+			panic("timer doesn't work through Interrupt-remapped IO-APIC");
+#endif
 		clear_IO_APIC_pin(apic1, pin1);
 		if (!no_pin1)
 			apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
@@ -2512,7 +2858,9 @@ static inline void __init check_timer(vo
 			    "through the IO-APIC - disabling NMI Watchdog!\n");
 		nmi_watchdog = NMI_NONE;
 	}
+#ifdef CONFIG_X86_32
 	timer_ack = 0;
+#endif
 
 	apic_printk(APIC_QUIET, KERN_INFO
 		    "...trying to set up timer as Virtual Wire IRQ...\n");
@@ -2570,17 +2918,25 @@ out:
 
 void __init setup_IO_APIC(void)
 {
+
+#ifdef CONFIG_X86_32
 	enable_IO_APIC();
+#else
+	/*
+	 * calling enable_IO_APIC() is moved to setup_local_APIC for BP
+	 */
+#endif
 
 	io_apic_irqs = ~PIC_IRQS;
 
-	printk("ENABLING IO-APIC IRQs\n");
-
-	/*
-	 * Set up IO-APIC IRQ routing.
-	 */
-	if (!acpi_ioapic)
-		setup_ioapic_ids_from_mpc();
+	apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
+        /*
+         * Set up IO-APIC IRQ routing.
+         */
+#ifdef CONFIG_X86_32
+        if (!acpi_ioapic)
+                setup_ioapic_ids_from_mpc();
+#endif
 	sync_Arb_IDs();
 	setup_IO_APIC_irqs();
 	init_IO_APIC_traps();
@@ -2588,15 +2944,15 @@ void __init setup_IO_APIC(void)
 }
 
 /*
- *	Called after all the initialization is done. If we didnt find any
- *	APIC bugs then we can allow the modify fast path
+ *      Called after all the initialization is done. If we didnt find any
+ *      APIC bugs then we can allow the modify fast path
  */
 
 static int __init io_apic_bug_finalize(void)
 {
-	if (sis_apic_bug == -1)
-		sis_apic_bug = 0;
-	return 0;
+        if (sis_apic_bug == -1)
+                sis_apic_bug = 0;
+        return 0;
 }
 
 late_initcall(io_apic_bug_finalize);
@@ -2605,7 +2961,7 @@ struct sysfs_ioapic_data {
 	struct sys_device dev;
 	struct IO_APIC_route_entry entry[0];
 };
-static struct sysfs_ioapic_data *mp_ioapic_data[MAX_IO_APICS];
+static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
 
 static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
 {
@@ -2615,8 +2971,8 @@ static int ioapic_suspend(struct sys_dev
 
 	data = container_of(dev, struct sysfs_ioapic_data, dev);
 	entry = data->entry;
-	for (i = 0; i < nr_ioapic_registers[dev->id]; i++)
-		entry[i] = ioapic_read_entry(dev->id, i);
+	for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ )
+		*entry = ioapic_read_entry(dev->id, i);
 
 	return 0;
 }
@@ -2653,14 +3009,14 @@ static struct sysdev_class ioapic_sysdev
 
 static int __init ioapic_init_sysfs(void)
 {
-	struct sys_device *dev;
-	int i, size, error = 0;
+	struct sys_device * dev;
+	int i, size, error;
 
 	error = sysdev_class_register(&ioapic_sysdev_class);
 	if (error)
 		return error;
 
-	for (i = 0; i < nr_ioapics; i++) {
+	for (i = 0; i < nr_ioapics; i++ ) {
 		size = sizeof(struct sys_device) + nr_ioapic_registers[i]
 			* sizeof(struct IO_APIC_route_entry);
 		mp_ioapic_data[i] = kzalloc(size, GFP_KERNEL);
@@ -2691,12 +3047,12 @@ device_initcall(ioapic_init_sysfs);
 unsigned int create_irq(unsigned int irq_want)
 {
 	/* Allocate an unused irq */
-	unsigned int irq, new;
+	unsigned int irq;
+	unsigned int new;
 	unsigned long flags;
 	struct irq_cfg *cfg_new;
 
 #ifndef CONFIG_HAVE_SPARSE_IRQ
-	/* only can use bus/dev/fn.. when per_cpu vector is used */
 	irq_want = nr_irqs - 1;
 #endif
 
@@ -2729,6 +3085,9 @@ void destroy_irq(unsigned int irq)
 
 	dynamic_irq_cleanup(irq);
 
+#ifdef CONFIG_INTR_REMAP
+	free_irte(irq);
+#endif
 	spin_lock_irqsave(&vector_lock, flags);
 	__clear_irq_vector(irq);
 	spin_unlock_irqrestore(&vector_lock, flags);
@@ -2754,25 +3113,54 @@ static int msi_compose_msg(struct pci_de
 	cpus_and(tmp, cfg->domain, tmp);
 	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(cfg->vector);
-
+#ifdef CONFIG_INTR_REMAP
+	if (irq_remapped(irq)) {
+		struct irte irte;
+		int ir_index;
+		u16 sub_handle;
+
+		ir_index = map_irq_to_irte_handle(irq, &sub_handle);
+		BUG_ON(ir_index == -1);
+
+		memset (&irte, 0, sizeof(irte));
+
+		irte.present = 1;
+		irte.dst_mode = INT_DEST_MODE;
+		irte.trigger_mode = 0; /* edge */
+		irte.dlvry_mode = INT_DELIVERY_MODE;
+		irte.vector = cfg->vector;
+		irte.dest_id = IRTE_DEST(dest);
+
+		modify_irte(irq, &irte);
+
+		msg->address_hi = MSI_ADDR_BASE_HI;
+		msg->data = sub_handle;
+		msg->address_lo = MSI_ADDR_BASE_LO | MSI_ADDR_IR_EXT_INT |
+				  MSI_ADDR_IR_SHV |
+				  MSI_ADDR_IR_INDEX1(ir_index) |
+				  MSI_ADDR_IR_INDEX2(ir_index);
+	} else
+#endif
+	{
+		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(cfg->vector);
+	}
 	return err;
 }
 
@@ -2805,6 +3193,56 @@ static void set_msi_irq_affinity(unsigne
 	write_msi_msg(irq, &msg);
 	desc->affinity = mask;
 }
+
+#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 ir_set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	unsigned int dest;
+	cpumask_t tmp, cleanup_mask;
+	struct irte irte;
+
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (get_irte(irq, &irte))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	irte.vector = cfg->vector;
+	irte.dest_id = IRTE_DEST(dest);
+
+	/*
+	 * atomically update the IRTE with the new destination and vector.
+	 */
+	modify_irte(irq, &irte);
+
+	/*
+	 * After this point, all the interrupts will start arriving
+	 * at the new destination. So, time to cleanup the previous
+	 * vector allocation.
+	 */
+	if (cfg->move_in_progress) {
+		cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
+		cfg->move_cleanup_count = cpus_weight(cleanup_mask);
+		send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+		cfg->move_in_progress = 0;
+	}
+
+	desc->affinity = mask;
+}
+#endif
 #endif /* CONFIG_SMP */
 
 /*
@@ -2822,6 +3260,45 @@ static struct irq_chip msi_chip = {
 	.retrigger	= ioapic_retrigger_irq,
 };
 
+#ifdef CONFIG_INTR_REMAP
+static struct irq_chip msi_ir_chip = {
+	.name		= "IR-PCI-MSI",
+	.unmask		= unmask_msi_irq,
+	.mask		= mask_msi_irq,
+	.ack		= ack_x2apic_edge,
+#ifdef CONFIG_SMP
+	.set_affinity	= ir_set_msi_irq_affinity,
+#endif
+	.retrigger	= ioapic_retrigger_irq,
+};
+
+/*
+ * Map the PCI dev to the corresponding remapping hardware unit
+ * and allocate 'nvec' consecutive interrupt-remapping table entries
+ * in it.
+ */
+static int msi_alloc_irte(struct pci_dev *dev, int irq, int nvec)
+{
+	struct intel_iommu *iommu;
+	int index;
+
+	iommu = map_dev_to_ir(dev);
+	if (!iommu) {
+		printk(KERN_ERR
+		       "Unable to map PCI %s to iommu\n", pci_name(dev));
+		return -ENOENT;
+	}
+
+	index = alloc_irte(iommu, irq, nvec);
+	if (index < 0) {
+		printk(KERN_ERR
+		       "Unable to allocate %d IRTE for PCI %s\n", nvec,
+		        pci_name(dev));
+		return -ENOSPC;
+	}
+	return index;
+}
+#endif
 
 static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc, int irq)
 {
@@ -2835,7 +3312,17 @@ static int setup_msi_irq(struct pci_dev 
 	set_irq_msi(irq, desc);
 	write_msi_msg(irq, &msg);
 
-	set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
+#ifdef CONFIG_INTR_REMAP
+	if (irq_remapped(irq)) {
+		struct irq_desc *desc = irq_desc(irq);
+		/*
+		 * irq migration in process context
+		 */
+		desc->status |= IRQ_MOVE_PCNTXT;
+		set_irq_chip_and_handler_name(irq, &msi_ir_chip, handle_edge_irq, "edge");
+	} else
+#endif
+		set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge");
 
 	return 0;
 }
@@ -2854,58 +3341,162 @@ static unsigned int build_irq_for_pci_de
 
 int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
 {
-	int irq, ret;
-        unsigned int irq_want;
-
-        irq_want = build_irq_for_pci_dev(dev) + 0x100;
+	unsigned int irq;
+	int ret;
+	unsigned int irq_want;
 
-        irq = create_irq(irq_want);
+	irq_want = build_irq_for_pci_dev(dev) + 0x100;
 
+	irq = create_irq(irq_want);
 	if (irq == 0)
 		return -1;
 
+#ifdef CONFIG_INTR_REMAP
+	if (!intr_remapping_enabled)
+		goto no_ir;
+
+	ret = msi_alloc_irte(dev, irq, 1);
+	if (ret < 0)
+		goto error;
+no_ir:
+#endif
 	ret = setup_msi_irq(dev, desc, irq);
 	if (ret < 0) {
 		destroy_irq(irq);
 		return ret;
-        }
-
+	}
 	return 0;
+
+#ifdef CONFIG_INTR_REMAP
+error:
+	destroy_irq(irq);
+	return ret;
+#endif
 }
 
 int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
 {
-        unsigned int irq;
-        int ret, sub_handle;
-        struct msi_desc *desc;
-        unsigned int irq_want;
-
-        irq_want = build_irq_for_pci_dev(dev) + 0x100;
-        sub_handle = 0;
-        list_for_each_entry(desc, &dev->msi_list, list) {
-                irq = create_irq(irq_want--);
-                if (irq == 0)
-                        return -1;
-                ret = setup_msi_irq(dev, desc, irq);
-                if (ret < 0)
-                        goto error;
-                sub_handle++;
-        }
-        return 0;
+	unsigned int irq;
+	int ret, sub_handle;
+	struct msi_desc *desc;
+	unsigned int irq_want;
+
+#ifdef CONFIG_INTR_REMAP
+	struct intel_iommu *iommu = 0;
+	int index = 0;
+#endif
+
+	irq_want = build_irq_for_pci_dev(dev) + 0x100;
+	sub_handle = 0;
+	list_for_each_entry(desc, &dev->msi_list, list) {
+		irq = create_irq(irq_want--);
+		if (irq == 0)
+			return -1;
+#ifdef CONFIG_INTR_REMAP
+		if (!intr_remapping_enabled)
+			goto no_ir;
+
+		if (!sub_handle) {
+			/*
+			 * allocate the consecutive block of IRTE's
+			 * for 'nvec'
+			 */
+			index = msi_alloc_irte(dev, irq, nvec);
+			if (index < 0) {
+				ret = index;
+				goto error;
+			}
+		} else {
+			iommu = map_dev_to_ir(dev);
+			if (!iommu) {
+				ret = -ENOENT;
+				goto error;
+			}
+			/*
+			 * setup the mapping between the irq and the IRTE
+			 * base index, the sub_handle pointing to the
+			 * appropriate interrupt remap table entry.
+			 */
+			set_irte_irq(irq, iommu, index, sub_handle);
+		}
+no_ir:
+#endif
+		ret = setup_msi_irq(dev, desc, irq);
+		if (ret < 0)
+			goto error;
+		sub_handle++;
+	}
+	return 0;
 
 error:
-        destroy_irq(irq);
-        return ret;
+	destroy_irq(irq);
+	return ret;
 }
 
-
 void arch_teardown_msi_irq(unsigned int irq)
 {
 	destroy_irq(irq);
 }
 
-#endif /* CONFIG_PCI_MSI */
+#ifdef CONFIG_DMAR
+#ifdef CONFIG_SMP
+static void dmar_msi_set_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+{
+	struct irq_cfg *cfg;
+	struct msi_msg msg;
+	unsigned int dest;
+	cpumask_t tmp;
 
+	cpus_and(tmp, mask, cpu_online_map);
+	if (cpus_empty(tmp))
+		return;
+
+	if (assign_irq_vector(irq, mask))
+		return;
+
+	cfg = irq_cfg(irq);
+	cpus_and(tmp, cfg->domain, mask);
+	dest = cpu_mask_to_apicid(tmp);
+
+	dmar_msi_read(irq, &msg);
+
+	msg.data &= ~MSI_DATA_VECTOR_MASK;
+	msg.data |= MSI_DATA_VECTOR(cfg->vector);
+	msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
+	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
+
+	dmar_msi_write(irq, &msg);
+	desc->affinity = mask;
+}
+#endif /* CONFIG_SMP */
+
+struct irq_chip dmar_msi_type = {
+	.name = "DMAR_MSI",
+	.unmask = dmar_msi_unmask,
+	.mask = dmar_msi_mask,
+	.ack = ack_apic_edge,
+#ifdef CONFIG_SMP
+	.set_affinity = dmar_msi_set_affinity,
+#endif
+	.retrigger = ioapic_retrigger_irq,
+};
+
+int arch_setup_dmar_msi(unsigned int irq)
+{
+	int ret;
+	struct msi_msg msg;
+
+	ret = msi_compose_msg(NULL, irq, &msg);
+	if (ret < 0)
+		return ret;
+	dmar_msi_write(irq, &msg);
+	set_irq_chip_and_handler_name(irq, &dmar_msi_type, handle_edge_irq,
+		"edge");
+	return 0;
+}
+#endif
+
+#endif /* CONFIG_PCI_MSI */
 /*
  * Hypertransport interrupt support
  */
@@ -2968,7 +3559,7 @@ int arch_setup_ht_irq(unsigned int irq, 
 
 	tmp = TARGET_CPUS;
 	err = assign_irq_vector(irq, tmp);
-	if ( !err) {
+	if (!err) {
 		struct ht_irq_msg msg;
 		unsigned dest;
 
@@ -3001,11 +3592,12 @@ int arch_setup_ht_irq(unsigned int irq, 
 #endif /* CONFIG_HT_IRQ */
 
 /* --------------------------------------------------------------------------
-			ACPI-based IOAPIC Configuration
+                          ACPI-based IOAPIC Configuration
    -------------------------------------------------------------------------- */
 
 #ifdef CONFIG_ACPI
 
+#ifdef CONFIG_X86_32
 int __init io_apic_get_unique_id(int ioapic, int apic_id)
 {
 	union IO_APIC_reg_00 reg_00;
@@ -3080,7 +3672,6 @@ int __init io_apic_get_unique_id(int ioa
 	return apic_id;
 }
 
-
 int __init io_apic_get_version(int ioapic)
 {
 	union IO_APIC_reg_01	reg_01;
@@ -3092,9 +3683,9 @@ int __init io_apic_get_version(int ioapi
 
 	return reg_01.bits.version;
 }
+#endif
 
-
-int __init io_apic_get_redir_entries(int ioapic)
+int __init io_apic_get_redir_entries (int ioapic)
 {
 	union IO_APIC_reg_01	reg_01;
 	unsigned long flags;
@@ -3107,10 +3698,10 @@ int __init io_apic_get_redir_entries(int
 }
 
 
-int io_apic_set_pci_routing(int ioapic, int pin, int irq, int triggering, int polarity)
+int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
 {
 	if (!IO_APIC_IRQ(irq)) {
-		printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
+		apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
 			ioapic);
 		return -EINVAL;
 	}
@@ -3126,6 +3717,7 @@ int io_apic_set_pci_routing(int ioapic, 
 	return 0;
 }
 
+
 int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
 {
 	int i;
@@ -3178,13 +3770,56 @@ void __init setup_ioapic_dest(void)
 				setup_IO_APIC_irq(ioapic, pin, irq,
 						  irq_trigger(irq_entry),
 						  irq_polarity(irq_entry));
+#ifdef CONFIG_INTR_REMAP
+			else if (intr_remapping_enabled) {
+				desc = irq_desc(irq);
+				set_ir_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
+			}
+#endif
 			else {
 				desc = irq_desc(irq);
 				set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
 			}
 		}
+	}
+}
+#endif
+
+#ifdef CONFIG_X86_64
+#define IOAPIC_RESOURCE_NAME_SIZE 11
+
+static struct resource *ioapic_resources;
+
+static struct resource * __init ioapic_setup_resources(void)
+{
+	unsigned long n;
+	struct resource *res;
+	char *mem;
+	int i;
+
+	if (nr_ioapics <= 0)
+		return NULL;
 
+	n = IOAPIC_RESOURCE_NAME_SIZE + sizeof(struct resource);
+	n *= nr_ioapics;
+
+	mem = alloc_bootmem(n);
+	res = (void *)mem;
+
+	if (mem != NULL) {
+		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;
+		}
 	}
+
+	ioapic_resources = res;
+
+	return res;
 }
 #endif
 
@@ -3192,29 +3827,70 @@ void __init ioapic_init_mappings(void)
 {
 	unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
 	int i;
+#ifdef CONFIG_X86_64
+	struct resource *ioapic_res;
 
+	ioapic_res = ioapic_setup_resources();
+#endif
 	for (i = 0; i < nr_ioapics; i++) {
 		if (smp_found_config) {
 			ioapic_phys = mp_ioapics[i].mp_apicaddr;
-			if (!ioapic_phys) {
-				printk(KERN_ERR
-				       "WARNING: bogus zero IO-APIC "
-				       "address found in MPTABLE, "
-				       "disabling IO/APIC support!\n");
-				smp_found_config = 0;
-				skip_ioapic_setup = 1;
-				goto fake_ioapic_page;
-			}
+#ifdef CONFIG_X86_32
+                        if (!ioapic_phys) {
+                                printk(KERN_ERR
+                                       "WARNING: bogus zero IO-APIC "
+                                       "address found in MPTABLE, "
+                                       "disabling IO/APIC support!\n");
+                                smp_found_config = 0;
+                                skip_ioapic_setup = 1;
+                                goto fake_ioapic_page;
+                        }
+#endif
 		} else {
+#ifdef CONFIG_X86_32
 fake_ioapic_page:
+#endif
 			ioapic_phys = (unsigned long)
-				      alloc_bootmem_pages(PAGE_SIZE);
+				alloc_bootmem_pages(PAGE_SIZE);
 			ioapic_phys = __pa(ioapic_phys);
 		}
 		set_fixmap_nocache(idx, ioapic_phys);
-		printk(KERN_DEBUG "mapped IOAPIC to %08lx (%08lx)\n",
-		       __fix_to_virt(idx), ioapic_phys);
+		apic_printk(APIC_VERBOSE,
+			    "mapped IOAPIC to %08lx (%08lx)\n",
+			    __fix_to_virt(idx), ioapic_phys);
 		idx++;
+
+#ifdef CONFIG_X86_64
+		if (ioapic_res != NULL) {
+			ioapic_res->start = ioapic_phys;
+			ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
+			ioapic_res++;
+		}
+#endif
 	}
 }
 
+#ifdef CONFIG_X86_64
+static int __init ioapic_insert_resources(void)
+{
+	int i;
+	struct resource *r = ioapic_resources;
+
+	if (!r) {
+		printk(KERN_ERR
+		       "IO APIC resources could be not be allocated.\n");
+		return -1;
+	}
+
+	for (i = 0; i < nr_ioapics; i++) {
+		insert_resource(&iomem_resource, r);
+		r++;
+	}
+
+	return 0;
+}
+
+/* Insert the IO APIC resources after PCI initialization has occured to handle
+ * IO APICS that are mapped in on a BAR in PCI space. */
+late_initcall(ioapic_insert_resources);
+#endif

[PATCH] x86: io_apic.c, build fix

[Ingo Molnar]

>From 306bb46f92d326c5afae3fa04860bdccb879b6b8 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Sat, 16 Aug 2008 10:18:31 +0200
Subject: [PATCH] x86: io_apic.c, build fix

fix:

 arch/x86/kernel/io_apic.c: In function 'MPBIOS_trigger':
 arch/x86/kernel/io_apic.c:1105: error: implicit declaration of function 'default_MCA_trigger'

the MCA defines got lost in the unification.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/kernel/io_apic.c |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c
index 2043cd8..929429f 100644
--- a/arch/x86/kernel/io_apic.c
+++ b/arch/x86/kernel/io_apic.c
@@ -1025,6 +1025,12 @@ static int EISA_ELCR(unsigned int irq)
 #define default_EISA_trigger(idx)       (EISA_ELCR(mp_irqs[idx].mp_srcbusirq))
 #define default_EISA_polarity(idx)      default_ISA_polarity(idx)
 
+/* 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)	default_ISA_polarity(idx)
+
 
 static int MPBIOS_polarity(int idx)
 {

Re: [PATCH] x86: io_apic.c, build fix

[Yinghai Lu]

On Sat, Aug 16, 2008 at 1:22 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
> From 306bb46f92d326c5afae3fa04860bdccb879b6b8 Mon Sep 17 00:00:00 2001
> From: Ingo Molnar <mingo@elte.hu>
> Date: Sat, 16 Aug 2008 10:18:31 +0200
> Subject: [PATCH] x86: io_apic.c, build fix
>
> fix:
>
>  arch/x86/kernel/io_apic.c: In function 'MPBIOS_trigger':
>  arch/x86/kernel/io_apic.c:1105: error: implicit declaration of function 'default_MCA_trigger'
>
> the MCA defines got lost in the unification.
>
> Signed-off-by: Ingo Molnar <mingo@elte.hu>
> ---
>  arch/x86/kernel/io_apic.c |    6 ++++++
>  1 files changed, 6 insertions(+), 0 deletions(-)
>
> diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c
> index 2043cd8..929429f 100644
> --- a/arch/x86/kernel/io_apic.c
> +++ b/arch/x86/kernel/io_apic.c
> @@ -1025,6 +1025,12 @@ static int EISA_ELCR(unsigned int irq)
>  #define default_EISA_trigger(idx)       (EISA_ELCR(mp_irqs[idx].mp_srcbusirq))
>  #define default_EISA_polarity(idx)      default_ISA_polarity(idx)
>
> +/* 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)      default_ISA_polarity(idx)
> +
>
>  static int MPBIOS_polarity(int idx)
>  {
>

ah. EISA got too copies

/* ISA interrupts are always polarity zero edge triggered,
 * when listed as conforming in the MP table. */

#define default_ISA_trigger(idx)        (0)
#define default_ISA_polarity(idx)       (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].mp_srcbusirq))
#define default_EISA_polarity(idx)      default_ISA_polarity(idx)

/* PCI interrupts are always polarity one level triggered,
 * when listed as conforming in the MP table. */

#define default_PCI_trigger(idx)        (1)
#define default_PCI_polarity(idx)       (1)

/* 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].mp_srcbusirq))
#define default_EISA_polarity(idx)      default_ISA_polarity(idx)


need to remove the  extra one.

YH

[PATCH] x86: apic - unify lapic_resume - fix

[Yinghai Lu]

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c
index 1b0baf5..7fb7dca 100644
--- a/arch/x86/kernel/apic_64.c
+++ b/arch/x86/kernel/apic_64.c
@@ -1448,6 +1448,7 @@ static int lapic_resume(struct sys_device *dev)
 		enable_x2apic();
 	else
 #endif
+	{
 		/*
 		 * Make sure the APICBASE points to the right address
 		 *
@@ -1458,6 +1459,7 @@ static int lapic_resume(struct sys_device *dev)
 		l &= ~MSR_IA32_APICBASE_BASE;
 		l |= MSR_IA32_APICBASE_ENABLE | mp_lapic_addr;
 		wrmsr(MSR_IA32_APICBASE, l, h);
+	}
 
 	apic_write(APIC_LVTERR, ERROR_APIC_VECTOR | APIC_LVT_MASKED);
 	apic_write(APIC_ID, apic_pm_state.apic_id);

[PATCH 1/2] x86: make HAVE_SPARSE_IRQ support selectable

[Yinghai Lu]

Ingo said sparse_irq is some intrusive. need to make it selectable

to make it simple, remove irq_desc as parameter in some functions.
(ack, eoi, set_affinity).
may need to make member if irq_chip to take irq_desc, or struct irq later.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/Kconfig              |    3 -
 arch/x86/Kconfig          |   12 ++++
 arch/x86/kernel/io_apic.c |  124 ++++++++++++++++++++++++++++++----------------
 arch/x86/kernel/irq_32.c  |    2 
 arch/x86/kernel/irq_64.c  |    4 -
 include/linux/irq.h       |   12 ----
 kernel/irq/chip.c         |   26 ---------
 kernel/irq/handle.c       |   34 +-----------
 kernel/irq/manage.c       |    8 --
 kernel/irq/migration.c    |    8 --
 10 files changed, 103 insertions(+), 130 deletions(-)

Index: linux-2.6/arch/Kconfig
===================================================================
--- linux-2.6.orig/arch/Kconfig
+++ linux-2.6/arch/Kconfig
@@ -106,6 +106,3 @@ config HAVE_CLK
 config HAVE_DYN_ARRAY
 	def_bool n
 
-config HAVE_SPARSE_IRQ
-	def_bool n
-
Index: linux-2.6/arch/x86/Kconfig
===================================================================
--- linux-2.6.orig/arch/x86/Kconfig
+++ linux-2.6/arch/x86/Kconfig
@@ -34,7 +34,6 @@ config X86
 	select HAVE_GENERIC_DMA_COHERENT if X86_32
 	select HAVE_EFFICIENT_UNALIGNED_ACCESS
 	select HAVE_DYN_ARRAY
-	select HAVE_SPARSE_IRQ
 
 config ARCH_DEFCONFIG
 	string
@@ -241,6 +240,17 @@ config SMP
 
 	  If you don't know what to do here, say N.
 
+config HAVE_SPARSE_IRQ
+	bool "Support sparse irq numbering"
+	depends on PCI_MSI || HT_IRQ
+	default y
+	help
+	  This enables support for sparse irq, esp for msi/msi-x. the irq
+	  number will be bus/dev/fn + 12bit. You may need if you have lots of
+	  cards supports msi-x installed.
+
+	  If you don't know what to do here, say Y.
+
 config X86_FIND_SMP_CONFIG
 	def_bool y
 	depends on X86_MPPARSE || X86_VOYAGER
Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic.c
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -107,7 +107,9 @@ struct irq_cfg;
 struct irq_pin_list;
 struct irq_cfg {
 	unsigned int irq;
+#ifdef CONFIG_HAVE_SPARSE_IRQ
 	struct irq_cfg *next;
+#endif
 	struct irq_pin_list *irq_2_pin;
 	cpumask_t domain;
 	cpumask_t old_domain;
@@ -137,20 +139,6 @@ static struct irq_cfg irq_cfg_legacy[] _
 };
 
 static struct irq_cfg irq_cfg_init = { .irq =  -1U, };
-/* need to be biger than size of irq_cfg_legacy */
-static int nr_irq_cfg = 32;
-
-static int __init parse_nr_irq_cfg(char *arg)
-{
-	if (arg) {
-		nr_irq_cfg = simple_strtoul(arg, NULL, 0);
-		if (nr_irq_cfg < 32)
-			nr_irq_cfg = 32;
-	}
-	return 0;
-}
-
-early_param("nr_irq_cfg", parse_nr_irq_cfg);
 
 static void init_one_irq_cfg(struct irq_cfg *cfg)
 {
@@ -158,7 +146,9 @@ static void init_one_irq_cfg(struct irq_
 }
 
 static struct irq_cfg *irq_cfgx;
+#ifdef CONFIG_HAVE_SPARSE_IRQ
 static struct irq_cfg *irq_cfgx_free;
+#endif
 static void __init init_work(void *data)
 {
 	struct dyn_array *da = data;
@@ -174,15 +164,34 @@ static void __init init_work(void *data)
 	for (i = legacy_count; i < *da->nr; i++)
 		init_one_irq_cfg(&cfg[i]);
 
+#ifdef CONFIG_HAVE_SPARSE_IRQ
 	for (i = 1; i < *da->nr; i++)
 		cfg[i-1].next = &cfg[i];
 
 	irq_cfgx_free = &irq_cfgx[legacy_count];
 	irq_cfgx[legacy_count - 1].next = NULL;
+#endif
 }
 
-#define for_each_irq_cfg(cfg)		\
-	for (cfg = irq_cfgx; cfg; cfg = cfg->next)
+#ifdef CONFIG_HAVE_SPARSE_IRQ
+/* need to be biger than size of irq_cfg_legacy */
+static int nr_irq_cfg = 32;
+
+static int __init parse_nr_irq_cfg(char *arg)
+{
+	if (arg) {
+		nr_irq_cfg = simple_strtoul(arg, NULL, 0);
+		if (nr_irq_cfg < 32)
+			nr_irq_cfg = 32;
+	}
+	return 0;
+}
+
+early_param("nr_irq_cfg", parse_nr_irq_cfg);
+
+#define for_each_irq_cfg(irqX, cfg)           \
+        for (cfg = irq_cfgx, irqX = cfg->irq; cfg; cfg = cfg->next, irqX = cfg ? cfg->irq : -1U)
+
 
 DEFINE_DYN_ARRAY(irq_cfgx, sizeof(struct irq_cfg), nr_irq_cfg, PAGE_SIZE, init_work);
 
@@ -273,7 +282,26 @@ static struct irq_cfg *irq_cfg_with_new(
 #endif
 	return cfg;
 }
+#else
 
+#define for_each_irq_cfg(irq, cfg)		\
+	for (irq = 0, cfg = &irq_cfgx[irq]; irq < nr_irqs; irq++, cfg = &irq_cfgx[irq])
+
+DEFINE_DYN_ARRAY(irq_cfgx, sizeof(struct irq_cfg), nr_irqs, PAGE_SIZE, init_work);
+
+struct irq_cfg *irq_cfg(unsigned int irq)
+{
+        if (irq < nr_irqs)
+                return &irq_cfgx[irq];
+
+        return NULL;
+}
+struct irq_cfg *irq_cfg_with_new(unsigned int irq)
+{
+        return irq_cfg(irq);
+}
+
+#endif
 /*
  * This is performance-critical, we want to do it O(1)
  *
@@ -506,12 +534,13 @@ static void __target_IO_APIC_irq(unsigne
 
 static int assign_irq_vector(int irq, cpumask_t mask);
 
-static void set_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
 {
 	struct irq_cfg *cfg;
 	unsigned long flags;
 	unsigned int dest;
 	cpumask_t tmp;
+	struct irq_desc *desc;
 
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
@@ -528,6 +557,7 @@ static void set_ioapic_affinity_irq(unsi
 	 */
 	dest = SET_APIC_LOGICAL_ID(dest);
 
+	desc = irq_desc(irq);
 	spin_lock_irqsave(&ioapic_lock, flags);
 	__target_IO_APIC_irq(irq, dest, cfg->vector);
 	desc->affinity = mask;
@@ -1289,11 +1319,10 @@ void __setup_vector_irq(int cpu)
 	struct irq_cfg *cfg;
 
 	/* Mark the inuse vectors */
-	for_each_irq_cfg(cfg) {
+	for_each_irq_cfg(irq, cfg) {
 		if (!cpu_isset(cpu, cfg->domain))
 			continue;
 		vector = cfg->vector;
-		irq = cfg->irq;
 		per_cpu(vector_irq, cpu)[vector] = irq;
 	}
 	/* Mark the free vectors */
@@ -1570,6 +1599,7 @@ __apicdebuginit(void) print_IO_APIC(void
 	union IO_APIC_reg_03 reg_03;
 	unsigned long flags;
 	struct irq_cfg *cfg;
+	unsigned int irq;
 
 	if (apic_verbosity == APIC_QUIET)
 		return;
@@ -1658,11 +1688,11 @@ __apicdebuginit(void) print_IO_APIC(void
 	}
 	}
 	printk(KERN_DEBUG "IRQ to pin mappings:\n");
-	for_each_irq_cfg(cfg) {
+	for_each_irq_cfg(irq, cfg) {
 		struct irq_pin_list *entry = cfg->irq_2_pin;
 		if (!entry)
 			continue;
-		printk(KERN_DEBUG "IRQ%d ", cfg->irq);
+		printk(KERN_DEBUG "IRQ%d ", irq);
 		for (;;) {
 			printk("-> %d:%d", entry->apic, entry->pin);
 			if (!entry->next)
@@ -2299,7 +2329,7 @@ static void ir_irq_migration(struct work
 				continue;
 			}
 
-			desc->chip->set_affinity(irq, desc, desc->pending_mask);
+			desc->chip->set_affinity(irq, desc->pending_mask);
 			spin_unlock_irqrestore(&desc->lock, flags);
 		}
 	}
@@ -2308,8 +2338,10 @@ static void ir_irq_migration(struct work
 /*
  * Migrates the IRQ destination in the process context.
  */
-static void set_ir_ioapic_affinity_irq(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+static void set_ir_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
 {
+	struct irq_desc *desc = irq_desc(irq);
+
 	if (desc->status & IRQ_LEVEL) {
 		desc->status |= IRQ_MOVE_PENDING;
 		desc->pending_mask = mask;
@@ -2381,19 +2413,21 @@ static void irq_complete_move(unsigned i
 static inline void irq_complete_move(unsigned int irq) {}
 #endif
 #ifdef CONFIG_INTR_REMAP
-static void ack_x2apic_level(unsigned int irq, struct irq_desc *desc)
+static void ack_x2apic_level(unsigned int irq)
 {
 	ack_x2APIC_irq();
 }
 
-static void ack_x2apic_edge(unsigned int irq, struct irq_desc *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)
+static void ack_apic_edge(unsigned int irq)
 {
+	struct irq_desc *desc = irq_desc(irq);
+
 	irq_complete_move(irq);
 	move_native_irq(irq, desc);
 	ack_APIC_irq();
@@ -2403,13 +2437,14 @@ static void ack_apic_edge(unsigned int i
 atomic_t irq_mis_count;
 #endif
 
-static void ack_apic_level(unsigned int irq, struct irq_desc *desc)
+static void ack_apic_level(unsigned int irq)
 {
 #ifdef CONFIG_X86_32
 	unsigned long v;
 	int i;
 #endif
 	int do_unmask_irq = 0;
+	struct irq_desc *desc = irq_desc(irq);
 
 	irq_complete_move(irq);
 #ifdef CONFIG_GENERIC_PENDING_IRQ
@@ -2540,8 +2575,7 @@ static inline void init_IO_APIC_traps(vo
 	 * Also, we've got to be careful not to trash gate
 	 * 0x80, because int 0x80 is hm, kind of importantish. ;)
 	 */
-	for_each_irq_cfg(cfg) {
-		irq = cfg->irq;
+	for_each_irq_cfg(irq, cfg) {
 		if (IO_APIC_IRQ(irq) && !cfg->vector) {
 			/*
 			 * Hmm.. We don't have an entry for this,
@@ -2579,7 +2613,7 @@ static void unmask_lapic_irq(unsigned in
 	apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
 }
 
-static void ack_lapic_irq(unsigned int irq, struct irq_desc *desc)
+static void ack_lapic_irq(unsigned int irq)
 {
 	ack_APIC_irq();
 }
@@ -3128,12 +3162,13 @@ static int msi_compose_msg(struct pci_de
 }
 
 #ifdef CONFIG_SMP
-static void set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+static void set_msi_irq_affinity(unsigned int irq, cpumask_t mask)
 {
 	struct irq_cfg *cfg;
 	struct msi_msg msg;
 	unsigned int dest;
 	cpumask_t tmp;
+	struct irq_desc *desc;
 
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
@@ -3154,6 +3189,7 @@ static void set_msi_irq_affinity(unsigne
 	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
 
 	write_msi_msg(irq, &msg);
+	desc = irq_desc(irq);
 	desc->affinity = mask;
 }
 
@@ -3162,12 +3198,13 @@ static void set_msi_irq_affinity(unsigne
  * Migrate the MSI irq to another cpumask. This migration is
  * done in the process context using interrupt-remapping hardware.
  */
-static void ir_set_msi_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+static void ir_set_msi_irq_affinity(unsigned int irq, cpumask_t mask)
 {
 	struct irq_cfg *cfg;
 	unsigned int dest;
 	cpumask_t tmp, cleanup_mask;
 	struct irte irte;
+	struct irq_desc *desc;
 
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
@@ -3203,6 +3240,7 @@ static void ir_set_msi_irq_affinity(unsi
 		cfg->move_in_progress = 0;
 	}
 
+	desc = irq_desc(irq);
 	desc->affinity = mask;
 }
 #endif
@@ -3403,12 +3441,13 @@ void arch_teardown_msi_irq(unsigned int
 
 #ifdef CONFIG_DMAR
 #ifdef CONFIG_SMP
-static void dmar_msi_set_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+static void dmar_msi_set_affinity(unsigned int irq, cpumask_t mask)
 {
 	struct irq_cfg *cfg;
 	struct msi_msg msg;
 	unsigned int dest;
 	cpumask_t tmp;
+	struct irq_desc *desc;
 
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
@@ -3429,6 +3468,8 @@ static void dmar_msi_set_affinity(unsign
 	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
 
 	dmar_msi_write(irq, &msg);
+
+	desc = irq_desc(irq);
 	desc->affinity = mask;
 }
 #endif /* CONFIG_SMP */
@@ -3481,11 +3522,12 @@ static void target_ht_irq(unsigned int i
 	write_ht_irq_msg(irq, &msg);
 }
 
-static void set_ht_irq_affinity(unsigned int irq, struct irq_desc *desc, cpumask_t mask)
+static void set_ht_irq_affinity(unsigned int irq, cpumask_t mask)
 {
 	struct irq_cfg *cfg;
 	unsigned int dest;
 	cpumask_t tmp;
+	struct irq_desc *desc;
 
 	cpus_and(tmp, mask, cpu_online_map);
 	if (cpus_empty(tmp))
@@ -3499,6 +3541,7 @@ static void set_ht_irq_affinity(unsigned
 	dest = cpu_mask_to_apicid(tmp);
 
 	target_ht_irq(irq, dest, cfg->vector);
+	desc = irq_desc(irq);
 	desc->affinity = mask;
 }
 #endif
@@ -3712,7 +3755,6 @@ void __init setup_ioapic_dest(void)
 {
 	int pin, ioapic, irq, irq_entry;
 	struct irq_cfg *cfg;
-	struct irq_desc *desc;
 
 	if (skip_ioapic_setup == 1)
 		return;
@@ -3734,15 +3776,11 @@ void __init setup_ioapic_dest(void)
 						  irq_trigger(irq_entry),
 						  irq_polarity(irq_entry));
 #ifdef CONFIG_INTR_REMAP
-			else if (intr_remapping_enabled) {
-				desc = irq_desc(irq);
-				set_ir_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
-			}
+			else if (intr_remapping_enabled)
+				set_ir_ioapic_affinity_irq(irq, TARGET_CPUS);
 #endif
-			else {
-				desc = irq_desc(irq);
-				set_ioapic_affinity_irq(irq, desc, TARGET_CPUS);
-			}
+			else
+				set_ioapic_affinity_irq(irq, TARGET_CPUS);
 		}
 	}
 }
Index: linux-2.6/arch/x86/kernel/irq_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irq_64.c
+++ linux-2.6/arch/x86/kernel/irq_64.c
@@ -74,7 +74,7 @@ int show_interrupts(struct seq_file *p,
 	struct irqaction * action;
 	unsigned long flags;
 	unsigned int entries;
-	struct irq_desc *desc;
+	struct irq_desc *desc = NULL;
 	int tail = 0;
 
 #ifdef CONFIG_HAVE_SPARSE_IRQ
@@ -275,7 +275,7 @@ void fixup_irqs(cpumask_t map)
 			desc->chip->mask(irq);
 
 		if (desc->chip->set_affinity)
-			desc->chip->set_affinity(irq, desc, mask);
+			desc->chip->set_affinity(irq, mask);
 		else if (!(warned++))
 			set_affinity = 0;
 
Index: linux-2.6/include/linux/irq.h
===================================================================
--- linux-2.6.orig/include/linux/irq.h
+++ linux-2.6/include/linux/irq.h
@@ -104,26 +104,14 @@ struct irq_chip {
 	void		(*enable)(unsigned int irq);
 	void		(*disable)(unsigned int irq);
 
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	void		(*ack)(unsigned int irq, struct irq_desc *desc);
-#else
 	void		(*ack)(unsigned int irq);
-#endif
 	void		(*mask)(unsigned int irq);
 	void		(*mask_ack)(unsigned int irq);
 	void		(*unmask)(unsigned int irq);
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	void		(*eoi)(unsigned int irq, struct irq_desc *desc);
-#else
 	void		(*eoi)(unsigned int irq);
-#endif
 
 	void		(*end)(unsigned int irq);
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	void		(*set_affinity)(unsigned int irq, struct irq_desc *desc, cpumask_t dest);
-#else
 	void		(*set_affinity)(unsigned int irq, cpumask_t dest);
-#endif
 	int		(*retrigger)(unsigned int irq);
 	int		(*set_type)(unsigned int irq, unsigned int flow_type);
 	int		(*set_wake)(unsigned int irq, unsigned int on);
Index: linux-2.6/kernel/irq/chip.c
===================================================================
--- linux-2.6.orig/kernel/irq/chip.c
+++ linux-2.6/kernel/irq/chip.c
@@ -299,11 +299,7 @@ static inline void mask_ack_irq(struct i
 		desc->chip->mask_ack(irq);
 	else {
 		desc->chip->mask(irq);
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-		desc->chip->ack(irq, desc);
-#else
 		desc->chip->ack(irq);
-#endif
 	}
 }
 
@@ -443,11 +439,7 @@ handle_fasteoi_irq(unsigned int irq, str
 	spin_lock(&desc->lock);
 	desc->status &= ~IRQ_INPROGRESS;
 out:
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	desc->chip->eoi(irq, desc);
-#else
 	desc->chip->eoi(irq);
-#endif
 
 	spin_unlock(&desc->lock);
 }
@@ -490,11 +482,7 @@ handle_edge_irq(unsigned int irq, struct
 	kstat_irqs_this_cpu(desc)++;
 
 	/* Start handling the irq */
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	desc->chip->ack(irq, desc);
-#else
 	desc->chip->ack(irq);
-#endif
 
 	/* Mark the IRQ currently in progress.*/
 	desc->status |= IRQ_INPROGRESS;
@@ -548,25 +536,15 @@ handle_percpu_irq(unsigned int irq, stru
 
 	kstat_irqs_this_cpu(desc)++;
 
-	if (desc->chip->ack) {
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-		desc->chip->ack(irq, desc);
-#else
+	if (desc->chip->ack)
 		desc->chip->ack(irq);
-#endif
-	}
 
 	action_ret = handle_IRQ_event(irq, desc->action);
 	if (!noirqdebug)
 		note_interrupt(irq, desc, action_ret);
 
-	if (desc->chip->eoi) {
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-		desc->chip->eoi(irq, desc);
-#else
+	if (desc->chip->eoi)
 		desc->chip->eoi(irq);
-#endif
-	}
 }
 
 void
Index: linux-2.6/kernel/irq/handle.c
===================================================================
--- linux-2.6.orig/kernel/irq/handle.c
+++ linux-2.6/kernel/irq/handle.c
@@ -285,13 +285,6 @@ struct irq_desc *irq_desc_with_new(unsig
  * What should we do if we get a hw irq event on an illegal vector?
  * Each architecture has to answer this themself.
  */
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-static void ack_bad(unsigned int irq, struct irq_desc *desc)
-{
-	print_irq_desc(irq, desc);
-	ack_bad_irq(irq);
-}
-#else
 static void ack_bad(unsigned int irq)
 {
 	struct irq_desc *desc;
@@ -300,7 +293,6 @@ static void ack_bad(unsigned int irq)
 	print_irq_desc(irq, desc);
 	ack_bad_irq(irq);
 }
-#endif
 
 /*
  * NOP functions
@@ -309,12 +301,6 @@ static void noop(unsigned int irq)
 {
 }
 
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-static void noop_desc(unsigned int irq, struct irq_desc *desc)
-{
-}
-#endif
-
 static unsigned int noop_ret(unsigned int irq)
 {
 	return 0;
@@ -343,11 +329,7 @@ struct irq_chip dummy_irq_chip = {
 	.shutdown	= noop,
 	.enable		= noop,
 	.disable	= noop,
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	.ack		= noop_desc,
-#else
 	.ack		= noop,
-#endif
 	.mask		= noop,
 	.unmask		= noop,
 	.end		= noop,
@@ -415,13 +397,9 @@ unsigned int __do_IRQ(unsigned int irq,
 		/*
 		 * No locking required for CPU-local interrupts:
 		 */
-		if (desc->chip->ack) {
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-			desc->chip->ack(irq, desc);
-#else
+		if (desc->chip->ack)
 			desc->chip->ack(irq);
-#endif
-		}
+
 		if (likely(!(desc->status & IRQ_DISABLED))) {
 			action_ret = handle_IRQ_event(irq, desc->action);
 			if (!noirqdebug)
@@ -432,13 +410,9 @@ unsigned int __do_IRQ(unsigned int irq,
 	}
 
 	spin_lock(&desc->lock);
-	if (desc->chip->ack) {
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-		desc->chip->ack(irq, desc);
-#else
+	if (desc->chip->ack)
 		desc->chip->ack(irq);
-#endif
-	}
+
 	/*
 	 * REPLAY is when Linux resends an IRQ that was dropped earlier
 	 * WAITING is used by probe to mark irqs that are being tested
Index: linux-2.6/kernel/irq/manage.c
===================================================================
--- linux-2.6.orig/kernel/irq/manage.c
+++ linux-2.6/kernel/irq/manage.c
@@ -91,11 +91,7 @@ int irq_set_affinity(unsigned int irq, c
 		unsigned long flags;
 
 		spin_lock_irqsave(&desc->lock, flags);
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-		desc->chip->set_affinity(irq, desc, cpumask);
-#else
 		desc->chip->set_affinity(irq, cpumask);
-#endif
 		spin_unlock_irqrestore(&desc->lock, flags);
 	} else
 		set_pending_irq(irq, cpumask);
@@ -122,11 +118,7 @@ int irq_select_affinity(unsigned int irq
 
 	desc = irq_desc(irq);
 	desc->affinity = mask;
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-	desc->chip->set_affinity(irq, desc, mask);
-#else
 	desc->chip->set_affinity(irq, mask);
-#endif
 
 	return 0;
 }
Index: linux-2.6/kernel/irq/migration.c
===================================================================
--- linux-2.6.orig/kernel/irq/migration.c
+++ linux-2.6/kernel/irq/migration.c
@@ -51,13 +51,9 @@ void move_masked_irq(int irq, struct irq
 	 * For correct operation this depends on the caller
 	 * masking the irqs.
 	 */
-	if (likely(!cpus_empty(tmp))) {
-#ifdef CONFIG_HAVE_SPARSE_IRQ
-		desc->chip->set_affinity(irq, desc, tmp);
-#else
+	if (likely(!cpus_empty(tmp)))
 		desc->chip->set_affinity(irq, tmp);
-#endif
-	}
+
 	cpus_clear(desc->pending_mask);
 }
 
Index: linux-2.6/arch/x86/kernel/irq_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irq_32.c
+++ linux-2.6/arch/x86/kernel/irq_32.c
@@ -269,7 +269,7 @@ int show_interrupts(struct seq_file *p,
 	struct irqaction * action;
 	unsigned long flags;
 	unsigned int entries;
-	struct irq_desc *desc;
+	struct irq_desc *desc = NULL;
 	int tail = 0;
 
 #ifdef CONFIG_HAVE_SPARSE_IRQ
@@ -434,7 +434,7 @@ void fixup_irqs(cpumask_t map)
 			mask = map;
 		}
 		if (desc->chip->set_affinity)
-			desc->chip->set_affinity(irq, desc, mask);
+			desc->chip->set_affinity(irq, mask);
 		else if (desc->action && !(warned++))
 			printk("Cannot set affinity for irq %i\n", irq);
 	}

[PATCH 2/2] irq: rename irq_desc() to to_irq_desc()

[Yinghai Lu]

So could revert back to use irq_desc[] for other arch except x86

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/alpha/kernel/irq.c                     |   21 +--
 arch/alpha/kernel/irq_alpha.c               |    7 -
 arch/alpha/kernel/irq_i8259.c               |    8 -
 arch/alpha/kernel/irq_pyxis.c               |    9 -
 arch/alpha/kernel/irq_srm.c                 |    9 -
 arch/alpha/kernel/sys_alcor.c               |    9 -
 arch/alpha/kernel/sys_cabriolet.c           |    8 -
 arch/alpha/kernel/sys_dp264.c               |    9 -
 arch/alpha/kernel/sys_eb64p.c               |    8 -
 arch/alpha/kernel/sys_eiger.c               |    7 -
 arch/alpha/kernel/sys_jensen.c              |   50 ++++-----
 arch/alpha/kernel/sys_marvel.c              |   20 +--
 arch/alpha/kernel/sys_mikasa.c              |    8 -
 arch/alpha/kernel/sys_noritake.c            |    8 -
 arch/alpha/kernel/sys_rawhide.c             |    8 -
 arch/alpha/kernel/sys_rx164.c               |    8 -
 arch/alpha/kernel/sys_sable.c               |    8 -
 arch/alpha/kernel/sys_takara.c              |    8 -
 arch/alpha/kernel/sys_titan.c               |    8 -
 arch/alpha/kernel/sys_wildfire.c            |   19 +--
 arch/arm/common/gic.c                       |    2 
 arch/arm/common/it8152.c                    |    2 
 arch/arm/common/locomo.c                    |   16 +--
 arch/arm/common/sa1111.c                    |    4 
 arch/arm/kernel/ecard.c                     |    4 
 arch/arm/kernel/irq.c                       |    8 -
 arch/arm/mach-at91/gpio.c                   |    4 
 arch/arm/mach-davinci/gpio.c                |   13 --
 arch/arm/mach-ep93xx/core.c                 |   12 +-
 arch/arm/mach-footbridge/isa-irq.c          |    2 
 arch/arm/mach-h720x/common.c                |    2 
 arch/arm/mach-h720x/cpu-h7202.c             |    2 
 arch/arm/mach-imx/irq.c                     |    2 
 arch/arm/mach-integrator/integrator_cp.c    |    2 
 arch/arm/mach-iop13xx/msi.c                 |    2 
 arch/arm/mach-ixp2000/core.c                |    4 
 arch/arm/mach-ixp2000/ixdp2x00.c            |    2 
 arch/arm/mach-ixp2000/ixdp2x01.c            |    2 
 arch/arm/mach-ixp23xx/core.c                |    2 
 arch/arm/mach-ixp23xx/ixdp2351.c            |    4 
 arch/arm/mach-lh7a40x/common.h              |    2 
 arch/arm/mach-netx/generic.c                |    2 
 arch/arm/mach-ns9xxx/board-a9m9750dev.c     |    2 
 arch/arm/mach-omap1/fpga.c                  |    2 
 arch/arm/mach-orion5x/irq.c                 |   12 +-
 arch/arm/mach-pxa/gpio.c                    |    2 
 arch/arm/mach-pxa/lpd270.c                  |    2 
 arch/arm/mach-pxa/lubbock.c                 |    2 
 arch/arm/mach-pxa/mainstone.c               |    2 
 arch/arm/mach-pxa/pcm990-baseboard.c        |    2 
 arch/arm/mach-s3c2410/bast-irq.c            |    6 -
 arch/arm/mach-s3c2412/irq.c                 |    4 
 arch/arm/mach-s3c2440/irq.c                 |    4 
 arch/arm/mach-s3c2443/irq.c                 |    4 
 arch/arm/mach-sa1100/h3600.c                |   14 +-
 arch/arm/mach-sa1100/irq.c                  |    2 
 arch/arm/mach-sa1100/neponset.c             |    6 -
 arch/arm/mach-versatile/core.c              |    2 
 arch/arm/oprofile/op_model_mpcore.c         |    2 
 arch/arm/plat-mxc/gpio.c                    |    7 -
 arch/arm/plat-omap/gpio.c                   |   12 --
 arch/arm/plat-orion/irq.c                   |    2 
 arch/arm/plat-s3c24xx/irq.c                 |   14 +-
 arch/arm/plat-s3c24xx/s3c244x-irq.c         |    4 
 arch/avr32/mach-at32ap/extint.c             |    7 -
 arch/avr32/mach-at32ap/intc.c               |    2 
 arch/avr32/mach-at32ap/pio.c                |    2 
 arch/blackfin/kernel/irqchip.c              |   12 --
 arch/blackfin/kernel/traps.c                |    8 -
 arch/blackfin/mach-bf537/boards/cm_bf537.c  |    2 
 arch/blackfin/mach-bf537/boards/stamp.c     |    2 
 arch/blackfin/mach-bf561/boards/cm_bf561.c  |    2 
 arch/blackfin/mach-common/ints-priority.c   |    8 -
 arch/cris/arch-v10/kernel/irq.c             |    2 
 arch/cris/arch-v32/kernel/irq.c             |    6 -
 arch/cris/kernel/irq.c                      |   12 --
 arch/frv/kernel/irq.c                       |   10 -
 arch/h8300/kernel/irq.c                     |   24 +---
 arch/ia64/hp/sim/hpsim_irq.c                |    2 
 arch/ia64/include/asm/hw_irq.h              |    2 
 arch/ia64/kernel/iosapic.c                  |   19 +--
 arch/ia64/kernel/irq.c                      |   16 +--
 arch/ia64/kernel/irq_ia64.c                 |   12 +-
 arch/ia64/kernel/mca.c                      |    2 
 arch/ia64/kernel/msi_ia64.c                 |    4 
 arch/ia64/kernel/smpboot.c                  |    2 
 arch/ia64/sn/kernel/irq.c                   |   11 --
 arch/ia64/sn/kernel/msi_sn.c                |    4 
 arch/m32r/kernel/irq.c                      |   12 --
 arch/m32r/platforms/m32104ut/setup.c        |   37 +++---
 arch/m32r/platforms/m32700ut/setup.c        |  133 +++++++++++-------------
 arch/m32r/platforms/mappi/setup.c           |   84 ++++++---------
 arch/m32r/platforms/mappi2/setup.c          |  103 +++++++------------
 arch/m32r/platforms/mappi3/setup.c          |  103 +++++++------------
 arch/m32r/platforms/oaks32r/setup.c         |   62 ++++-------
 arch/m32r/platforms/opsput/setup.c          |  149 +++++++++++-----------------
 arch/m32r/platforms/usrv/setup.c            |   92 ++++++-----------
 arch/m68knommu/kernel/irq.c                 |   18 +--
 arch/mips/au1000/common/irq.c               |    2 
 arch/mips/dec/ioasic-irq.c                  |    2 
 arch/mips/emma2rh/markeins/irq_markeins.c   |    2 
 arch/mips/kernel/irq-gic.c                  |    2 
 arch/mips/kernel/irq-msc01.c                |    2 
 arch/mips/kernel/irq.c                      |   10 -
 arch/mips/kernel/smtc.c                     |    2 
 arch/mips/mti-malta/malta-smtc.c            |    2 
 arch/mips/sgi-ip22/ip22-int.c               |    2 
 arch/mips/sgi-ip22/ip22-time.c              |    2 
 arch/mips/sgi-ip32/ip32-irq.c               |   10 -
 arch/mips/sibyte/bcm1480/irq.c              |    5 
 arch/mips/sibyte/bcm1480/smp.c              |    2 
 arch/mips/sibyte/sb1250/irq.c               |    5 
 arch/mips/sibyte/sb1250/smp.c               |    2 
 arch/mips/sni/a20r.c                        |    2 
 arch/mips/sni/pcimt.c                       |    2 
 arch/mips/sni/pcit.c                        |    2 
 arch/mips/sni/rm200.c                       |    2 
 arch/mips/vr41xx/common/icu.c               |   44 ++++----
 arch/mips/vr41xx/common/irq.c               |    2 
 arch/mn10300/kernel/irq.c                   |   13 --
 arch/parisc/kernel/irq.c                    |   41 +++----
 arch/powerpc/include/asm/irq.h              |    2 
 arch/powerpc/kernel/crash.c                 |    2 
 arch/powerpc/kernel/irq.c                   |   15 +-
 arch/powerpc/platforms/8xx/m8xx_setup.c     |    2 
 arch/powerpc/platforms/powermac/pic.c       |    4 
 arch/powerpc/platforms/pseries/eeh_driver.c |    2 
 arch/powerpc/platforms/pseries/xics.c       |    4 
 arch/powerpc/sysdev/cpm2_pic.c              |    5 
 arch/powerpc/sysdev/mpic.c                  |   16 +--
 arch/powerpc/sysdev/qe_lib/qe_ic.c          |    2 
 arch/sh/boards/mach-cayman/irq.c            |    4 
 arch/sh/boards/mach-dreamcast/irq.c         |    2 
 arch/sh/boards/mach-dreamcast/setup.c       |    2 
 arch/sh/boards/mach-microdev/irq.c          |    4 
 arch/sh/boards/mach-se/7206/irq.c           |    2 
 arch/sh/boards/mach-se/7343/irq.c           |    2 
 arch/sh/boards/mach-se/7722/irq.c           |    2 
 arch/sh/boards/mach-systemh/irq.c           |    4 
 arch/sh/cchips/hd6446x/hd64461.c            |    4 
 arch/sh/cchips/hd6446x/hd64465/setup.c      |    4 
 arch/sh/kernel/cpu/irq/imask.c              |    4 
 arch/sh/kernel/cpu/irq/intc-sh5.c           |    4 
 arch/sh/kernel/irq.c                        |   12 --
 arch/sparc64/kernel/irq.c                   |   41 +++----
 arch/sparc64/kernel/pci_msi.c               |    2 
 arch/sparc64/kernel/time.c                  |    2 
 arch/um/kernel/irq.c                        |   31 ++---
 arch/x86/kernel/io_apic.c                   |   32 +++---
 arch/x86/kernel/irq_32.c                    |    4 
 arch/x86/kernel/irq_64.c                    |    4 
 arch/x86/kernel/irqinit_32.c                |    2 
 arch/x86/kernel/irqinit_64.c                |    2 
 arch/x86/kernel/visws_quirks.c              |    8 -
 arch/x86/mach-voyager/voyager_smp.c         |    4 
 arch/xtensa/kernel/irq.c                    |   12 --
 drivers/char/random.c                       |    4 
 drivers/gpio/gpiolib.c                      |    2 
 drivers/mfd/asic3.c                         |    4 
 drivers/mfd/htc-egpio.c                     |    2 
 drivers/parisc/dino.c                       |    6 -
 drivers/parisc/eisa.c                       |    6 -
 drivers/parisc/gsc.c                        |    8 -
 drivers/parisc/iosapic.c                    |    2 
 drivers/parisc/superio.c                    |    4 
 drivers/pci/intr_remapping.c                |    4 
 drivers/pcmcia/hd64465_ss.c                 |    4 
 drivers/xen/events.c                        |    6 -
 include/asm-mips/irq.h                      |    2 
 include/linux/irq.h                         |   26 ++--
 kernel/irq/autoprobe.c                      |   10 -
 kernel/irq/chip.c                           |   28 ++---
 kernel/irq/handle.c                         |   24 ++--
 kernel/irq/manage.c                         |   26 ++--
 kernel/irq/migration.c                      |    2 
 kernel/irq/proc.c                           |   12 +-
 kernel/irq/resend.c                         |    2 
 kernel/irq/spurious.c                       |    4 
 178 files changed, 877 insertions(+), 1129 deletions(-)

Index: linux-2.6/arch/alpha/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq.c
+++ linux-2.6/arch/alpha/kernel/irq.c
@@ -46,19 +46,16 @@ int irq_select_affinity(unsigned int irq
 {
 	static int last_cpu;
 	int cpu = last_cpu + 1;
-	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
-
-	if (!desc->chip->set_affinity || irq_user_affinity[irq])
+	if (!irq_desc[irq].chip->set_affinity || irq_user_affinity[irq])
 		return 1;
 
 	while (!cpu_possible(cpu) || !cpu_isset(cpu, irq_default_affinity))
 		cpu = (cpu < (NR_CPUS-1) ? cpu + 1 : 0);
 	last_cpu = cpu;
 
-	desc->affinity = cpumask_of_cpu(cpu);
-	desc->chip->set_affinity(irq, cpumask_of_cpu(cpu));
+	irq_desc[irq].affinity = cpumask_of_cpu(cpu);
+	irq_desc[irq].chip->set_affinity(irq, cpumask_of_cpu(cpu));
 	return 0;
 }
 #endif /* CONFIG_SMP */
@@ -83,10 +80,8 @@ show_interrupts(struct seq_file *p, void
 #endif
 
 	if (irq < ACTUAL_NR_IRQS) {
-		struct irq_desc *desc = irq_desc(irq);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[irq].lock, flags);
+		action = irq_desc[irq].action;
 		if (!action) 
 			goto unlock;
 		seq_printf(p, "%3d: ", irq);
@@ -96,7 +91,7 @@ show_interrupts(struct seq_file *p, void
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(irq, j));
 #endif
-		seq_printf(p, " %14s", desc->chip->typename);
+		seq_printf(p, " %14s", irq_desc[irq].chip->typename);
 		seq_printf(p, "  %c%s",
 			(action->flags & IRQF_DISABLED)?'+':' ',
 			action->name);
@@ -109,7 +104,7 @@ show_interrupts(struct seq_file *p, void
 
 		seq_putc(p, '\n');
 unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[irq].lock, flags);
 	} else if (irq == ACTUAL_NR_IRQS) {
 #ifdef CONFIG_SMP
 		seq_puts(p, "IPI: ");
@@ -161,6 +156,6 @@ handle_irq(int irq)
 	 * at IPL 0.
 	 */
 	local_irq_disable();
-	__do_IRQ(irq, irq_desc(irq));
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 }
Index: linux-2.6/arch/alpha/kernel/irq_alpha.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq_alpha.c
+++ linux-2.6/arch/alpha/kernel/irq_alpha.c
@@ -64,7 +64,7 @@ do_entInt(unsigned long type, unsigned l
 		smp_percpu_timer_interrupt(regs);
 		cpu = smp_processor_id();
 		if (cpu != boot_cpuid) {
-			irq_desc(RTC_IRQ)->kstat_irqs[cpu]++;
+			irq_desc[RTC_IRQ].kstat_irqs[cpu]++;
 		} else {
 			handle_irq(RTC_IRQ);
 		}
@@ -240,9 +240,8 @@ static struct hw_interrupt_type rtc_irq_
 void __init
 init_rtc_irq(void)
 {
-	struct irq_desc *desc = irq_desc(RTC_IRQ);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &rtc_irq_type;
+	irq_desc[RTC_IRQ].status = IRQ_DISABLED;
+	irq_desc[RTC_IRQ].chip = &rtc_irq_type;
 	setup_irq(RTC_IRQ, &timer_irqaction);
 }
 
Index: linux-2.6/arch/alpha/kernel/irq_i8259.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq_i8259.c
+++ linux-2.6/arch/alpha/kernel/irq_i8259.c
@@ -79,7 +79,7 @@ i8259a_startup_irq(unsigned int irq)
 void
 i8259a_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		i8259a_enable_irq(irq);
 }
 
@@ -107,10 +107,8 @@ init_i8259a_irqs(void)
 	outb(0xff, 0xA1);	/* mask all of 8259A-2 */
 
 	for (i = 0; i < 16; i++) {
-		struct irq_desc *desc;
-		desc = irq_desc(i);
-		desc->status = IRQ_DISABLED;
-		desc->chip = &i8259a_irq_type;
+		irq_desc[i].status = IRQ_DISABLED;
+		irq_desc[i].chip = &i8259a_irq_type;
 	}
 
 	setup_irq(2, &cascade);
Index: linux-2.6/arch/alpha/kernel/irq_pyxis.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq_pyxis.c
+++ linux-2.6/arch/alpha/kernel/irq_pyxis.c
@@ -50,7 +50,7 @@ pyxis_startup_irq(unsigned int irq)
 static void
 pyxis_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		pyxis_enable_irq(irq);
 }
 
@@ -117,13 +117,10 @@ init_pyxis_irqs(unsigned long ignore_mas
 	*(vuip) CIA_IACK_SC;
 
 	for (i = 16; i < 48; ++i) {
-		struct irq_desc *desc;
-
 		if ((ignore_mask >> i) & 1)
 			continue;
-		desc = irq_desc(i);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &pyxis_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &pyxis_irq_type;
 	}
 
 	setup_irq(16+7, &isa_cascade_irqaction);
Index: linux-2.6/arch/alpha/kernel/irq_srm.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq_srm.c
+++ linux-2.6/arch/alpha/kernel/irq_srm.c
@@ -43,7 +43,7 @@ srm_startup_irq(unsigned int irq)
 static void
 srm_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		srm_enable_irq(irq);
 }
 
@@ -64,13 +64,10 @@ init_srm_irqs(long max, unsigned long ig
 	long i;
 
 	for (i = 16; i < max; ++i) {
-		struct irq_desc *desc;
-
 		if (i < 64 && ((ignore_mask >> i) & 1))
 			continue;
-		desc = irq_desc(i);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &srm_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &srm_irq_type;
 	}
 }
 
Index: linux-2.6/arch/alpha/kernel/sys_alcor.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_alcor.c
+++ linux-2.6/arch/alpha/kernel/sys_alcor.c
@@ -85,7 +85,7 @@ alcor_isa_mask_and_ack_irq(unsigned int
 static void
 alcor_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		alcor_enable_irq(irq);
 }
 
@@ -137,16 +137,13 @@ alcor_init_irq(void)
 	*(vuip)GRU_INT_CLEAR = 0; mb();			/* all clear */
 
 	for (i = 16; i < 48; ++i) {
-		struct irq_desc *desc;
-
 		/* On Alcor, at least, lines 20..30 are not connected
 		   and can generate spurious interrupts if we turn them
 		   on while IRQ probing.  */
 		if (i >= 16+20 && i <= 16+30)
 			continue;
-		desc = irq_desc(i);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &alcor_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &alcor_irq_type;
 	}
 	i8259a_irq_type.ack = alcor_isa_mask_and_ack_irq;
 
Index: linux-2.6/arch/alpha/kernel/sys_cabriolet.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_cabriolet.c
+++ linux-2.6/arch/alpha/kernel/sys_cabriolet.c
@@ -67,7 +67,7 @@ cabriolet_startup_irq(unsigned int irq)
 static void
 cabriolet_end_irq(unsigned int irq)
 { 
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		cabriolet_enable_irq(irq);
 }
 
@@ -122,10 +122,8 @@ common_init_irq(void (*srm_dev_int)(unsi
 		outb(0xff, 0x806);
 
 		for (i = 16; i < 35; ++i) {
-			struct irq_desc *desc = irq_desc(i);
-
-			desc->status = IRQ_DISABLED | IRQ_LEVEL;
-			desc->chip = &cabriolet_irq_type;
+			irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+			irq_desc[i].chip = &cabriolet_irq_type;
 		}
 	}
 
Index: linux-2.6/arch/alpha/kernel/sys_dp264.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_dp264.c
+++ linux-2.6/arch/alpha/kernel/sys_dp264.c
@@ -125,7 +125,7 @@ dp264_startup_irq(unsigned int irq)
 static void
 dp264_end_irq(unsigned int irq)
 { 
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		dp264_enable_irq(irq);
 }
 
@@ -157,7 +157,7 @@ clipper_startup_irq(unsigned int irq)
 static void
 clipper_end_irq(unsigned int irq)
 { 
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		clipper_enable_irq(irq);
 }
 
@@ -298,9 +298,8 @@ init_tsunami_irqs(struct hw_interrupt_ty
 {
 	long i;
 	for (i = imin; i <= imax; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = ops;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = ops;
 	}
 }
 
Index: linux-2.6/arch/alpha/kernel/sys_eb64p.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_eb64p.c
+++ linux-2.6/arch/alpha/kernel/sys_eb64p.c
@@ -65,7 +65,7 @@ eb64p_startup_irq(unsigned int irq)
 static void
 eb64p_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		eb64p_enable_irq(irq);
 }
 
@@ -135,10 +135,8 @@ eb64p_init_irq(void)
 	init_i8259a_irqs();
 
 	for (i = 16; i < 32; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &eb64p_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &eb64p_irq_type;
 	}		
 
 	common_init_isa_dma();
Index: linux-2.6/arch/alpha/kernel/sys_eiger.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_eiger.c
+++ linux-2.6/arch/alpha/kernel/sys_eiger.c
@@ -76,7 +76,7 @@ eiger_startup_irq(unsigned int irq)
 static void
 eiger_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		eiger_enable_irq(irq);
 }
 
@@ -153,9 +153,8 @@ eiger_init_irq(void)
 	init_i8259a_irqs();
 
 	for (i = 16; i < 128; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &eiger_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &eiger_irq_type;
 	}
 }
 
Index: linux-2.6/arch/alpha/kernel/sys_jensen.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_jensen.c
+++ linux-2.6/arch/alpha/kernel/sys_jensen.c
@@ -68,16 +68,13 @@ jensen_local_startup(unsigned int irq)
 	/* the parport is really hw IRQ 1, silly Jensen.  */
 	if (irq == 7)
 		i8259a_startup_irq(1);
-	else {
-		struct irq_desc *desc = irq_desc(irq);
-
+	else
 		/*
 		 * For all true local interrupts, set the flag that prevents
 		 * the IPL from being dropped during handler processing.
 		 */
-		if (desc->action)
-			desc->action->flags |= IRQF_DISABLED;
-	}
+		if (irq_desc[irq].action)
+			irq_desc[irq].action->flags |= IRQF_DISABLED;
 	return 0;
 }
 
@@ -161,21 +158,22 @@ jensen_device_interrupt(unsigned long ve
 	}
 
 	/* If there is no handler yet... */
-	if (irq_desc(irq)->action == NULL) {
-		/* If it is a local interrupt that cannot be masked... */
-		if (vector >= 0x900) {
-			/* Clear keyboard/mouse state */
-			inb(0x64);
-			inb(0x60);
-			/* Reset serial ports */
-			inb(0x3fa);
-			inb(0x2fa);
-			outb(0x0c, 0x3fc);
-			outb(0x0c, 0x2fc);
-			/* Clear NMI */
-			outb(0, 0x61);
-			outb(0, 0x461);
-		}
+	if (irq_desc[irq].action == NULL) {
+	    /* If it is a local interrupt that cannot be masked... */
+	    if (vector >= 0x900)
+	    {
+	        /* Clear keyboard/mouse state */
+	    	inb(0x64);
+		inb(0x60);
+		/* Reset serial ports */
+		inb(0x3fa);
+		inb(0x2fa);
+		outb(0x0c, 0x3fc);
+		outb(0x0c, 0x2fc);
+		/* Clear NMI */
+		outb(0,0x61);
+		outb(0,0x461);
+	    }
 	}
 
 #if 0
@@ -208,11 +206,11 @@ jensen_init_irq(void)
 {
 	init_i8259a_irqs();
 
-	irq_desc(1)->chip = &jensen_local_irq_type;
-	irq_desc(4)->chip = &jensen_local_irq_type;
-	irq_desc(3)->chip = &jensen_local_irq_type;
-	irq_desc(7)->chip = &jensen_local_irq_type;
-	irq_desc(9)->chip = &jensen_local_irq_type;
+	irq_desc[1].chip = &jensen_local_irq_type;
+	irq_desc[4].chip = &jensen_local_irq_type;
+	irq_desc[3].chip = &jensen_local_irq_type;
+	irq_desc[7].chip = &jensen_local_irq_type;
+	irq_desc[9].chip = &jensen_local_irq_type;
 
 	common_init_isa_dma();
 }
Index: linux-2.6/arch/alpha/kernel/sys_marvel.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_marvel.c
+++ linux-2.6/arch/alpha/kernel/sys_marvel.c
@@ -152,7 +152,7 @@ io7_startup_irq(unsigned int irq)
 static void
 io7_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		io7_enable_irq(irq);
 }
 
@@ -303,10 +303,8 @@ init_io7_irqs(struct io7 *io7,
 
 	/* Set up the lsi irqs.  */
 	for (i = 0; i < 128; ++i) {
-		struct irq_desc *desc = irq_desc(base + i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = lsi_ops;
+		irq_desc[base + i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[base + i].chip = lsi_ops;
 	}
 
 	/* Disable the implemented irqs in hardware.  */
@@ -319,10 +317,8 @@ init_io7_irqs(struct io7 *io7,
 
 	/* Set up the msi irqs.  */
 	for (i = 128; i < (128 + 512); ++i) {
-		struct irq_desc *desc = irq_desc(base + i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = msi_ops;
+		irq_desc[base + i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[base + i].chip = msi_ops;
 	}
 
 	for (i = 0; i < 16; ++i)
@@ -339,10 +335,8 @@ marvel_init_irq(void)
 
 	/* Reserve the legacy irqs.  */
 	for (i = 0; i < 16; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED;
-		desc->chip = &marvel_legacy_irq_type;
+		irq_desc[i].status = IRQ_DISABLED;
+		irq_desc[i].chip = &marvel_legacy_irq_type;
 	}
 
 	/* Init the io7 irqs.  */
Index: linux-2.6/arch/alpha/kernel/sys_mikasa.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_mikasa.c
+++ linux-2.6/arch/alpha/kernel/sys_mikasa.c
@@ -64,7 +64,7 @@ mikasa_startup_irq(unsigned int irq)
 static void
 mikasa_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		mikasa_enable_irq(irq);
 }
 
@@ -115,10 +115,8 @@ mikasa_init_irq(void)
 	mikasa_update_irq_hw(0);
 
 	for (i = 16; i < 32; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &mikasa_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &mikasa_irq_type;
 	}
 
 	init_i8259a_irqs();
Index: linux-2.6/arch/alpha/kernel/sys_noritake.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_noritake.c
+++ linux-2.6/arch/alpha/kernel/sys_noritake.c
@@ -69,7 +69,7 @@ noritake_startup_irq(unsigned int irq)
 static void
 noritake_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+        if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
                 noritake_enable_irq(irq);
 }
 
@@ -144,10 +144,8 @@ noritake_init_irq(void)
 	outw(0, 0x54c);
 
 	for (i = 16; i < 48; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &noritake_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &noritake_irq_type;
 	}
 
 	init_i8259a_irqs();
Index: linux-2.6/arch/alpha/kernel/sys_rawhide.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_rawhide.c
+++ linux-2.6/arch/alpha/kernel/sys_rawhide.c
@@ -131,7 +131,7 @@ rawhide_startup_irq(unsigned int irq)
 static void
 rawhide_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		rawhide_enable_irq(irq);
 }
 
@@ -194,10 +194,8 @@ rawhide_init_irq(void)
 	}
 
 	for (i = 16; i < 128; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &rawhide_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &rawhide_irq_type;
 	}
 
 	init_i8259a_irqs();
Index: linux-2.6/arch/alpha/kernel/sys_rx164.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_rx164.c
+++ linux-2.6/arch/alpha/kernel/sys_rx164.c
@@ -68,7 +68,7 @@ rx164_startup_irq(unsigned int irq)
 static void
 rx164_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		rx164_enable_irq(irq);
 }
 
@@ -116,10 +116,8 @@ rx164_init_irq(void)
 
 	rx164_update_irq_hw(0);
 	for (i = 16; i < 40; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &rx164_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &rx164_irq_type;
 	}
 
 	init_i8259a_irqs();
Index: linux-2.6/arch/alpha/kernel/sys_sable.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_sable.c
+++ linux-2.6/arch/alpha/kernel/sys_sable.c
@@ -484,7 +484,7 @@ sable_lynx_startup_irq(unsigned int irq)
 static void
 sable_lynx_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		sable_lynx_enable_irq(irq);
 }
 
@@ -535,10 +535,8 @@ sable_lynx_init_irq(int nr_irqs)
 	long i;
 
 	for (i = 0; i < nr_irqs; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &sable_lynx_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &sable_lynx_irq_type;
 	}
 
 	common_init_isa_dma();
Index: linux-2.6/arch/alpha/kernel/sys_takara.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_takara.c
+++ linux-2.6/arch/alpha/kernel/sys_takara.c
@@ -70,7 +70,7 @@ takara_startup_irq(unsigned int irq)
 static void
 takara_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		takara_enable_irq(irq);
 }
 
@@ -153,10 +153,8 @@ takara_init_irq(void)
 		takara_update_irq_hw(i, -1);
 
 	for (i = 16; i < 128; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &takara_irq_type;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = &takara_irq_type;
 	}
 
 	common_init_isa_dma();
Index: linux-2.6/arch/alpha/kernel/sys_titan.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_titan.c
+++ linux-2.6/arch/alpha/kernel/sys_titan.c
@@ -139,7 +139,7 @@ titan_startup_irq(unsigned int irq)
 static void
 titan_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		titan_enable_irq(irq);
 }
 
@@ -187,10 +187,8 @@ init_titan_irqs(struct hw_interrupt_type
 {
 	long i;
 	for (i = imin; i <= imax; ++i) {
-		struct irq_desc *desc = irq_desc(i);
-
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = ops;
+		irq_desc[i].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i].chip = ops;
 	}
 }
 
Index: linux-2.6/arch/alpha/kernel/sys_wildfire.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/sys_wildfire.c
+++ linux-2.6/arch/alpha/kernel/sys_wildfire.c
@@ -150,10 +150,10 @@ static void
 wildfire_end_irq(unsigned int irq)
 { 
 #if 0
-	if (!irq_desc(irq)->action)
+	if (!irq_desc[irq].action)
 		printk("got irq %d\n", irq);
 #endif
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		wildfire_enable_irq(irq);
 }
 
@@ -196,20 +196,17 @@ wildfire_init_irq_per_pca(int qbbno, int
 #endif
 
 	for (i = 0; i < 16; ++i) {
-		struct irq_desc *desc;
 		if (i == 2)
 			continue;
-		desc = irq_desc(i+irq_bias);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &wildfire_irq_type;
+		irq_desc[i+irq_bias].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i+irq_bias].chip = &wildfire_irq_type;
 	}
 
-	irq_desc(36+irq_bias)->status = IRQ_DISABLED | IRQ_LEVEL;
-	irq_desc(36+irq_bias)->chip = &wildfire_irq_type;
+	irq_desc[36+irq_bias].status = IRQ_DISABLED | IRQ_LEVEL;
+	irq_desc[36+irq_bias].chip = &wildfire_irq_type;
 	for (i = 40; i < 64; ++i) {
-		struct irq_desc *desc = irq_desc(i+irq_bias);
-		desc->status = IRQ_DISABLED | IRQ_LEVEL;
-		desc->chip = &wildfire_irq_type;
+		irq_desc[i+irq_bias].status = IRQ_DISABLED | IRQ_LEVEL;
+		irq_desc[i+irq_bias].chip = &wildfire_irq_type;
 	}
 
 	setup_irq(32+irq_bias, &isa_enable);	
Index: linux-2.6/arch/arm/common/gic.c
===================================================================
--- linux-2.6.orig/arch/arm/common/gic.c
+++ linux-2.6/arch/arm/common/gic.c
@@ -117,7 +117,7 @@ static void gic_set_cpu(unsigned int irq
 	u32 val;
 
 	spin_lock(&irq_controller_lock);
-	irq_desc(irq)->cpu = cpu;
+	irq_desc[irq].cpu = cpu;
 	val = readl(reg) & ~(0xff << shift);
 	val |= 1 << (cpu + shift);
 	writel(val, reg);
Index: linux-2.6/arch/arm/common/it8152.c
===================================================================
--- linux-2.6.orig/arch/arm/common/it8152.c
+++ linux-2.6/arch/arm/common/it8152.c
@@ -70,7 +70,7 @@ static inline void it8152_irq(int irq)
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	desc_handle_irq(irq, desc);
 }
 
Index: linux-2.6/arch/arm/common/locomo.c
===================================================================
--- linux-2.6.orig/arch/arm/common/locomo.c
+++ linux-2.6/arch/arm/common/locomo.c
@@ -181,8 +181,8 @@ static void locomo_handler(unsigned int
 	if (req) {
 		/* generate the next interrupt(s) */
 		irq = LOCOMO_IRQ_START;
-		d = irq_desc(irq);
-		for (i = 0; i <= 3; i++, irq++, d = irq_desc(irq)) {
+		d = irq_desc + irq;
+		for (i = 0; i <= 3; i++, d++, irq++) {
 			if (req & (0x0100 << i)) {
 				desc_handle_irq(irq, d);
 			}
@@ -226,7 +226,7 @@ static void locomo_key_handler(unsigned
 	void __iomem *mapbase = get_irq_chip_data(irq);
 
 	if (locomo_readl(mapbase + LOCOMO_KEYBOARD + LOCOMO_KIC) & 0x0001) {
-		d = irq_desc(LOCOMO_IRQ_KEY_START);
+		d = irq_desc + LOCOMO_IRQ_KEY_START;
 		desc_handle_irq(LOCOMO_IRQ_KEY_START, d);
 	}
 }
@@ -277,8 +277,8 @@ static void locomo_gpio_handler(unsigned
 
 	if (req) {
 		irq = LOCOMO_IRQ_GPIO_START;
-		d = irq_desc(LOCOMO_IRQ_GPIO_START);
-		for (i = 0; i <= 15; i++, irq++, d = irq_desc(irq)) {
+		d = irq_desc + LOCOMO_IRQ_GPIO_START;
+		for (i = 0; i <= 15; i++, irq++, d++) {
 			if (req & (0x0001 << i)) {
 				desc_handle_irq(irq, d);
 			}
@@ -365,7 +365,7 @@ static void locomo_lt_handler(unsigned i
 	void __iomem *mapbase = get_irq_chip_data(irq);
 
 	if (locomo_readl(mapbase + LOCOMO_LTINT) & 0x0001) {
-		d = irq_desc(LOCOMO_IRQ_LT_START);
+		d = irq_desc + LOCOMO_IRQ_LT_START;
 		desc_handle_irq(LOCOMO_IRQ_LT_START, d);
 	}
 }
@@ -413,9 +413,9 @@ static void locomo_spi_handler(unsigned
 	req = locomo_readl(mapbase + LOCOMO_SPI + LOCOMO_SPIIR) & 0x000F;
 	if (req) {
 		irq = LOCOMO_IRQ_SPI_START;
-		d = irq_desc(irq);
+		d = irq_desc + irq;
 
-		for (i = 0; i <= 3; i++, irq++, d = irq_desc(irq)) {
+		for (i = 0; i <= 3; i++, irq++, d++) {
 			if (req & (0x0001 << i)) {
 				desc_handle_irq(irq, d);
 			}
Index: linux-2.6/arch/arm/common/sa1111.c
===================================================================
--- linux-2.6.orig/arch/arm/common/sa1111.c
+++ linux-2.6/arch/arm/common/sa1111.c
@@ -170,11 +170,11 @@ sa1111_irq_handler(unsigned int irq, str
 
 	for (i = IRQ_SA1111_START; stat0; i++, stat0 >>= 1)
 		if (stat0 & 1)
-			handle_edge_irq(i, irq_desc(i));
+			handle_edge_irq(i, irq_desc + i);
 
 	for (i = IRQ_SA1111_START + 32; stat1; i++, stat1 >>= 1)
 		if (stat1 & 1)
-			handle_edge_irq(i, irq_desc(i));
+			handle_edge_irq(i, irq_desc + i);
 
 	/* For level-based interrupts */
 	desc->chip->unmask(irq);
Index: linux-2.6/arch/arm/kernel/ecard.c
===================================================================
--- linux-2.6.orig/arch/arm/kernel/ecard.c
+++ linux-2.6/arch/arm/kernel/ecard.c
@@ -587,7 +587,7 @@ ecard_irq_handler(unsigned int irq, stru
 			pending = ecard_default_ops.irqpending(ec);
 
 		if (pending) {
-			struct irq_desc *d = irq_desc(ec->irq);
+			struct irq_desc *d = irq_desc + ec->irq;
 			desc_handle_irq(ec->irq, d);
 			called ++;
 		}
@@ -622,7 +622,7 @@ ecard_irqexp_handler(unsigned int irq, s
 		ecard_t *ec = slot_to_ecard(slot);
 
 		if (ec->claimed) {
-			struct irq_desc *d = irq_desc(ec->irq);
+			struct irq_desc *d = irq_desc + ec->irq;
 			/*
 			 * this ugly code is so that we can operate a
 			 * prioritorising system:
Index: linux-2.6/arch/arm/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/kernel/irq.c
+++ linux-2.6/arch/arm/kernel/irq.c
@@ -112,7 +112,7 @@ static struct irq_desc bad_irq_desc = {
 asmlinkage void __exception asm_do_IRQ(unsigned int irq, struct pt_regs *regs)
 {
 	struct pt_regs *old_regs = set_irq_regs(regs);
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 
 	/*
 	 * Some hardware gives randomly wrong interrupts.  Rather
@@ -142,7 +142,7 @@ void set_irq_flags(unsigned int irq, uns
 		return;
 	}
 
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	spin_lock_irqsave(&desc->lock, flags);
 	desc->status |= IRQ_NOREQUEST | IRQ_NOPROBE | IRQ_NOAUTOEN;
 	if (iflags & IRQF_VALID)
@@ -159,7 +159,7 @@ void __init init_IRQ(void)
 	int irq;
 
 	for (irq = 0; irq < NR_IRQS; irq++)
-		irq_desc(irq)->status |= IRQ_NOREQUEST | IRQ_NOPROBE;
+		irq_desc[irq].status |= IRQ_NOREQUEST | IRQ_NOPROBE;
 
 #ifdef CONFIG_SMP
 	bad_irq_desc.affinity = CPU_MASK_ALL;
@@ -189,7 +189,7 @@ void migrate_irqs(void)
 	unsigned int i, cpu = smp_processor_id();
 
 	for (i = 0; i < NR_IRQS; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = irq_desc + i;
 
 		if (desc->cpu == cpu) {
 			unsigned int newcpu = any_online_cpu(desc->affinity);
Index: linux-2.6/arch/arm/mach-at91/gpio.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-at91/gpio.c
+++ linux-2.6/arch/arm/mach-at91/gpio.c
@@ -404,7 +404,7 @@ static void gpio_irq_handler(unsigned ir
 		}
 
 		pin = bank->chipbase;
-		gpio = irq_desc(pin);
+		gpio = &irq_desc[pin];
 
 		while (isr) {
 			if (isr & 1) {
@@ -515,7 +515,7 @@ void __init at91_gpio_irq_setup(void)
 		__raw_writel(~0, this->regbase + PIO_IDR);
 
 		for (i = 0, pin = this->chipbase; i < 32; i++, pin++) {
-			lockdep_set_class(&irq_desc(pin)->lock, &gpio_lock_class);
+			lockdep_set_class(&irq_desc[pin].lock, &gpio_lock_class);
 
 			/*
 			 * Can use the "simple" and not "edge" handler since it's
Index: linux-2.6/arch/arm/mach-davinci/gpio.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-davinci/gpio.c
+++ linux-2.6/arch/arm/mach-davinci/gpio.c
@@ -155,11 +155,10 @@ static void gpio_irq_enable(unsigned irq
 {
 	struct gpio_controller *__iomem g = get_irq_chip_data(irq);
 	u32 mask = __gpio_mask(irq_to_gpio(irq));
-	struct irq_desc *desc = irq_desc(irq);
 
-	if (desc->status & IRQ_TYPE_EDGE_FALLING)
+	if (irq_desc[irq].status & IRQ_TYPE_EDGE_FALLING)
 		__raw_writel(mask, &g->set_falling);
-	if (desc->status & IRQ_TYPE_EDGE_RISING)
+	if (irq_desc[irq].status & IRQ_TYPE_EDGE_RISING)
 		__raw_writel(mask, &g->set_rising);
 }
 
@@ -167,14 +166,12 @@ static int gpio_irq_type(unsigned irq, u
 {
 	struct gpio_controller *__iomem g = get_irq_chip_data(irq);
 	u32 mask = __gpio_mask(irq_to_gpio(irq));
-	struct irq_desc *desc;
 
 	if (trigger & ~(IRQ_TYPE_EDGE_FALLING | IRQ_TYPE_EDGE_RISING))
 		return -EINVAL;
 
-	desc = irq_desc(irq);
-	desc->status &= ~IRQ_TYPE_SENSE_MASK;
-	desc->status |= trigger;
+	irq_desc[irq].status &= ~IRQ_TYPE_SENSE_MASK;
+	irq_desc[irq].status |= trigger;
 
 	__raw_writel(mask, (trigger & IRQ_TYPE_EDGE_FALLING)
 		     ? &g->set_falling : &g->clr_falling);
@@ -218,7 +215,7 @@ gpio_irq_handler(unsigned irq, struct ir
 
 		/* now demux them to the right lowlevel handler */
 		n = (int)get_irq_data(irq);
-		gpio = irq_desc(n);
+		gpio = &irq_desc[n];
 		while (status) {
 			res = ffs(status);
 			n += res;
Index: linux-2.6/arch/arm/mach-ep93xx/core.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ep93xx/core.c
+++ linux-2.6/arch/arm/mach-ep93xx/core.c
@@ -192,7 +192,7 @@ static void ep93xx_gpio_ab_irq_handler(u
 	for (i = 0; i < 8; i++) {
 		if (status & (1 << i)) {
 			int gpio_irq = gpio_to_irq(EP93XX_GPIO_LINE_A(0)) + i;
-			desc = irq_desc(gpio_irq);
+			desc = irq_desc + gpio_irq;
 			desc_handle_irq(gpio_irq, desc);
 		}
 	}
@@ -201,7 +201,7 @@ static void ep93xx_gpio_ab_irq_handler(u
 	for (i = 0; i < 8; i++) {
 		if (status & (1 << i)) {
 			int gpio_irq = gpio_to_irq(EP93XX_GPIO_LINE_B(0)) + i;
-			desc = irq_desc(gpio_irq);
+			desc = irq_desc + gpio_irq;
 			desc_handle_irq(gpio_irq, desc);
 		}
 	}
@@ -217,7 +217,7 @@ static void ep93xx_gpio_f_irq_handler(un
 	int port_f_idx = ((irq + 1) & 7) ^ 4; /* {19..22,47..50} -> {0..7} */
 	int gpio_irq = gpio_to_irq(EP93XX_GPIO_LINE_F(0)) + port_f_idx;
 
-	desc_handle_irq(gpio_irq, irq_desc(gpio_irq));
+	desc_handle_irq(gpio_irq, irq_desc + gpio_irq);
 }
 
 static void ep93xx_gpio_irq_ack(unsigned int irq)
@@ -226,7 +226,7 @@ static void ep93xx_gpio_irq_ack(unsigned
 	int port = line >> 3;
 	int port_mask = 1 << (line & 7);
 
-	if ((irq_desc(irq)->status & IRQ_TYPE_SENSE_MASK) == IRQ_TYPE_EDGE_BOTH) {
+	if ((irq_desc[irq].status & IRQ_TYPE_SENSE_MASK) == IRQ_TYPE_EDGE_BOTH) {
 		gpio_int_type2[port] ^= port_mask; /* switch edge direction */
 		ep93xx_gpio_update_int_params(port);
 	}
@@ -240,7 +240,7 @@ static void ep93xx_gpio_irq_mask_ack(uns
 	int port = line >> 3;
 	int port_mask = 1 << (line & 7);
 
-	if ((irq_desc(irq)->status & IRQ_TYPE_SENSE_MASK) == IRQ_TYPE_EDGE_BOTH)
+	if ((irq_desc[irq].status & IRQ_TYPE_SENSE_MASK) == IRQ_TYPE_EDGE_BOTH)
 		gpio_int_type2[port] ^= port_mask; /* switch edge direction */
 
 	gpio_int_unmasked[port] &= ~port_mask;
@@ -275,7 +275,7 @@ static void ep93xx_gpio_irq_unmask(unsig
  */
 static int ep93xx_gpio_irq_type(unsigned int irq, unsigned int type)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	const int gpio = irq_to_gpio(irq);
 	const int port = gpio >> 3;
 	const int port_mask = 1 << (gpio & 7);
Index: linux-2.6/arch/arm/mach-footbridge/isa-irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-footbridge/isa-irq.c
+++ linux-2.6/arch/arm/mach-footbridge/isa-irq.c
@@ -94,7 +94,7 @@ isa_irq_handler(unsigned int irq, struct
 		return;
 	}
 
-	desc = irq_desc(isa_irq);
+	desc = irq_desc + isa_irq;
 	desc_handle_irq(isa_irq, desc);
 }
 
Index: linux-2.6/arch/arm/mach-h720x/common.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-h720x/common.c
+++ linux-2.6/arch/arm/mach-h720x/common.c
@@ -104,7 +104,7 @@ h720x_gpio_handler(unsigned int mask, un
                  struct irq_desc *desc)
 {
 	IRQDBG("%s irq: %d\n", __func__, irq);
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	while (mask) {
 		if (mask & 1) {
 			IRQDBG("handling irq %d\n", irq);
Index: linux-2.6/arch/arm/mach-h720x/cpu-h7202.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-h720x/cpu-h7202.c
+++ linux-2.6/arch/arm/mach-h720x/cpu-h7202.c
@@ -120,7 +120,7 @@ h7202_timerx_demux_handler(unsigned int
 
 	mask >>= 1;
 	irq = IRQ_TIMER1;
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	while (mask) {
 		if (mask & 1)
 			desc_handle_irq(irq, desc);
Index: linux-2.6/arch/arm/mach-imx/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-imx/irq.c
+++ linux-2.6/arch/arm/mach-imx/irq.c
@@ -182,7 +182,7 @@ static void
 imx_gpio_handler(unsigned int mask, unsigned int irq,
                  struct irq_desc *desc)
 {
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	while (mask) {
 		if (mask & 1) {
 			DEBUG_IRQ("handling irq %d\n", irq);
Index: linux-2.6/arch/arm/mach-integrator/integrator_cp.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-integrator/integrator_cp.c
+++ linux-2.6/arch/arm/mach-integrator/integrator_cp.c
@@ -217,7 +217,7 @@ sic_handle_irq(unsigned int irq, struct
 
 		irq += IRQ_SIC_START;
 
-		desc = irq_desc(irq);
+		desc = irq_desc + irq;
 		desc_handle_irq(irq, desc);
 	} while (status);
 }
Index: linux-2.6/arch/arm/mach-iop13xx/msi.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-iop13xx/msi.c
+++ linux-2.6/arch/arm/mach-iop13xx/msi.c
@@ -110,7 +110,7 @@ static void iop13xx_msi_handler(unsigned
 		do {
 			j = find_first_bit(&status, 32);
 			(write_imipr[i])(1 << j); /* write back to clear bit */
-			desc = irq_desc(IRQ_IOP13XX_MSI_0 + j + (32*i));
+			desc = irq_desc + IRQ_IOP13XX_MSI_0 + j + (32*i);
 			desc_handle_irq(IRQ_IOP13XX_MSI_0 + j + (32*i),	desc);
 			status = (read_imipr[i])();
 		} while (status);
Index: linux-2.6/arch/arm/mach-ixp2000/core.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ixp2000/core.c
+++ linux-2.6/arch/arm/mach-ixp2000/core.c
@@ -311,7 +311,7 @@ static void ixp2000_GPIO_irq_handler(uns
 		   
 	for (i = 0; i <= 7; i++) {
 		if (status & (1<<i)) {
-			desc = irq_desc(i + IRQ_IXP2000_GPIO0);
+			desc = irq_desc + i + IRQ_IXP2000_GPIO0;
 			desc_handle_irq(i + IRQ_IXP2000_GPIO0, desc);
 		}
 	}
@@ -404,7 +404,7 @@ static void ixp2000_err_irq_handler(unsi
 
 	for(i = 31; i >= 0; i--) {
 		if(status & (1 << i)) {
-			desc = irq_desc(IRQ_IXP2000_DRAM0_MIN_ERR + i);
+			desc = irq_desc + IRQ_IXP2000_DRAM0_MIN_ERR + i;
 			desc_handle_irq(IRQ_IXP2000_DRAM0_MIN_ERR + i, desc);
 		}
 	}
Index: linux-2.6/arch/arm/mach-ixp2000/ixdp2x00.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ixp2000/ixdp2x00.c
+++ linux-2.6/arch/arm/mach-ixp2000/ixdp2x00.c
@@ -131,7 +131,7 @@ static void ixdp2x00_irq_handler(unsigne
 		if(ex_interrupt & (1 << i))  {
 			struct irq_desc *cpld_desc;
 			int cpld_irq = IXP2000_BOARD_IRQ(0) + i;
-			cpld_desc = irq_desc(cpld_irq);
+			cpld_desc = irq_desc + cpld_irq;
 			desc_handle_irq(cpld_irq, cpld_desc);
 		}
 	}
Index: linux-2.6/arch/arm/mach-ixp2000/ixdp2x01.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ixp2000/ixdp2x01.c
+++ linux-2.6/arch/arm/mach-ixp2000/ixdp2x01.c
@@ -81,7 +81,7 @@ static void ixdp2x01_irq_handler(unsigne
 		if (ex_interrupt & (1 << i)) {
 			struct irq_desc *cpld_desc;
 			int cpld_irq = IXP2000_BOARD_IRQ(0) + i;
-			cpld_desc = irq_desc(cpld_irq);
+			cpld_desc = irq_desc + cpld_irq;
 			desc_handle_irq(cpld_irq, cpld_desc);
 		}
 	}
Index: linux-2.6/arch/arm/mach-ixp23xx/core.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ixp23xx/core.c
+++ linux-2.6/arch/arm/mach-ixp23xx/core.c
@@ -268,7 +268,7 @@ static void pci_handler(unsigned int irq
 		BUG();
 	}
 
-	int_desc = irq_desc(irqno);
+	int_desc = irq_desc + irqno;
 	desc_handle_irq(irqno, int_desc);
 
 	desc->chip->unmask(irq);
Index: linux-2.6/arch/arm/mach-ixp23xx/ixdp2351.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ixp23xx/ixdp2351.c
+++ linux-2.6/arch/arm/mach-ixp23xx/ixdp2351.c
@@ -71,7 +71,7 @@ static void ixdp2351_inta_handler(unsign
 			struct irq_desc *cpld_desc;
 			int cpld_irq =
 				IXP23XX_MACH_IRQ(IXDP2351_INTA_IRQ_BASE + i);
-			cpld_desc = irq_desc(cpld_irq);
+			cpld_desc = irq_desc + cpld_irq;
 			desc_handle_irq(cpld_irq, cpld_desc);
 		}
 	}
@@ -108,7 +108,7 @@ static void ixdp2351_intb_handler(unsign
 			struct irq_desc *cpld_desc;
 			int cpld_irq =
 				IXP23XX_MACH_IRQ(IXDP2351_INTB_IRQ_BASE + i);
-			cpld_desc = irq_desc(cpld_irq);
+			cpld_desc = irq_desc + cpld_irq;
 			desc_handle_irq(cpld_irq, cpld_desc);
 		}
 	}
Index: linux-2.6/arch/arm/mach-lh7a40x/common.h
===================================================================
--- linux-2.6.orig/arch/arm/mach-lh7a40x/common.h
+++ linux-2.6/arch/arm/mach-lh7a40x/common.h
@@ -15,4 +15,4 @@ extern void lh7a404_init_irq (void);
 extern void lh7a40x_clcd_init (void);
 extern void lh7a40x_init_board_irq (void);
 
-#define IRQ_DISPATCH(irq) desc_handle_irq((irq),(irq_desc(irq)))
+#define IRQ_DISPATCH(irq) desc_handle_irq((irq),(irq_desc + irq))
Index: linux-2.6/arch/arm/mach-netx/generic.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-netx/generic.c
+++ linux-2.6/arch/arm/mach-netx/generic.c
@@ -77,7 +77,7 @@ netx_hif_demux_handler(unsigned int irq_
 	stat = ((readl(NETX_DPMAS_INT_EN) &
 		readl(NETX_DPMAS_INT_STAT)) >> 24) & 0x1f;
 
-	desc = irq_desc(NETX_IRQ_HIF_CHAINED(0));
+	desc = irq_desc + NETX_IRQ_HIF_CHAINED(0);
 
 	while (stat) {
 		if (stat & 1) {
Index: linux-2.6/arch/arm/mach-ns9xxx/board-a9m9750dev.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ns9xxx/board-a9m9750dev.c
+++ linux-2.6/arch/arm/mach-ns9xxx/board-a9m9750dev.c
@@ -90,7 +90,7 @@ static void a9m9750dev_fpga_demux_handle
 
 		stat &= ~(1 << irqno);
 
-		fpgadesc = irq_desc(FPGA_IRQ(irqno));
+		fpgadesc = irq_desc + FPGA_IRQ(irqno);
 
 		desc_handle_irq(FPGA_IRQ(irqno), fpgadesc);
 	}
Index: linux-2.6/arch/arm/mach-omap1/fpga.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-omap1/fpga.c
+++ linux-2.6/arch/arm/mach-omap1/fpga.c
@@ -99,7 +99,7 @@ void innovator_fpga_IRQ_demux(unsigned i
 	     (fpga_irq < OMAP_FPGA_IRQ_END) && stat;
 	     fpga_irq++, stat >>= 1) {
 		if (stat & 1) {
-			d = irq_desc(fpga_irq);
+			d = irq_desc + fpga_irq;
 			desc_handle_irq(fpga_irq, d);
 		}
 	}
Index: linux-2.6/arch/arm/mach-orion5x/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-orion5x/irq.c
+++ linux-2.6/arch/arm/mach-orion5x/irq.c
@@ -47,7 +47,7 @@
 static void orion5x_gpio_irq_ack(u32 irq)
 {
 	int pin = irq_to_gpio(irq);
-	if (irq_desc(irq)->status & IRQ_LEVEL)
+	if (irq_desc[irq].status & IRQ_LEVEL)
 		/*
 		 * Mask bit for level interrupt
 		 */
@@ -62,7 +62,7 @@ static void orion5x_gpio_irq_ack(u32 irq
 static void orion5x_gpio_irq_mask(u32 irq)
 {
 	int pin = irq_to_gpio(irq);
-	if (irq_desc(irq)->status & IRQ_LEVEL)
+	if (irq_desc[irq].status & IRQ_LEVEL)
 		orion5x_clrbits(GPIO_LEVEL_MASK, 1 << pin);
 	else
 		orion5x_clrbits(GPIO_EDGE_MASK, 1 << pin);
@@ -71,7 +71,7 @@ static void orion5x_gpio_irq_mask(u32 ir
 static void orion5x_gpio_irq_unmask(u32 irq)
 {
 	int pin = irq_to_gpio(irq);
-	if (irq_desc(irq)->status & IRQ_LEVEL)
+	if (irq_desc[irq].status & IRQ_LEVEL)
 		orion5x_setbits(GPIO_LEVEL_MASK, 1 << pin);
 	else
 		orion5x_setbits(GPIO_EDGE_MASK, 1 << pin);
@@ -88,7 +88,7 @@ static int orion5x_gpio_set_irq_type(u32
 		return -EINVAL;
 	}
 
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 
 	switch (type) {
 	case IRQ_TYPE_LEVEL_HIGH:
@@ -155,7 +155,7 @@ static void orion5x_gpio_irq_handler(uns
 	for (pin = offs; pin < offs + 8; pin++) {
 		if (cause & (1 << pin)) {
 			irq = gpio_to_irq(pin);
-			desc = irq_desc(irq);
+			desc = irq_desc + irq;
 			if ((desc->status & IRQ_TYPE_SENSE_MASK) == IRQ_TYPE_EDGE_BOTH) {
 				/* Swap polarity (race with GPIO line) */
 				u32 polarity = readl(GPIO_IN_POL);
@@ -186,7 +186,7 @@ static void __init orion5x_init_gpio_irq
 	for (i = IRQ_ORION5X_GPIO_START; i < NR_IRQS; i++) {
 		set_irq_chip(i, &orion5x_gpio_irq_chip);
 		set_irq_handler(i, handle_level_irq);
-		desc = irq_desc(i);
+		desc = irq_desc + i;
 		desc->status |= IRQ_LEVEL;
 		set_irq_flags(i, IRQF_VALID);
 	}
Index: linux-2.6/arch/arm/mach-pxa/gpio.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-pxa/gpio.c
+++ linux-2.6/arch/arm/mach-pxa/gpio.c
@@ -275,7 +275,7 @@ static void pxa_gpio_demux_handler(unsig
 			loop = 1;
 
 			n = PXA_GPIO_IRQ_BASE + bit;
-			desc_handle_irq(n, irq_desc(n));
+			desc_handle_irq(n, irq_desc + n);
 
 			bit = find_next_bit(gedr, GEDR_BITS, bit + 1);
 		}
Index: linux-2.6/arch/arm/mach-pxa/lpd270.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-pxa/lpd270.c
+++ linux-2.6/arch/arm/mach-pxa/lpd270.c
@@ -88,7 +88,7 @@ static void lpd270_irq_handler(unsigned
 		GEDR(0) = GPIO_bit(0);  /* clear useless edge notification */
 		if (likely(pending)) {
 			irq = LPD270_IRQ(0) + __ffs(pending);
-			desc = irq_desc(irq);
+			desc = irq_desc + irq;
 			desc_handle_irq(irq, desc);
 
 			pending = __raw_readw(LPD270_INT_STATUS) &
Index: linux-2.6/arch/arm/mach-pxa/lubbock.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-pxa/lubbock.c
+++ linux-2.6/arch/arm/mach-pxa/lubbock.c
@@ -131,7 +131,7 @@ static void lubbock_irq_handler(unsigned
 		GEDR(0) = GPIO_bit(0);	/* clear our parent irq */
 		if (likely(pending)) {
 			irq = LUBBOCK_IRQ(0) + __ffs(pending);
-			desc = irq_desc(irq);
+			desc = irq_desc + irq;
 			desc_handle_irq(irq, desc);
 		}
 		pending = LUB_IRQ_SET_CLR & lubbock_irq_enabled;
Index: linux-2.6/arch/arm/mach-pxa/mainstone.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-pxa/mainstone.c
+++ linux-2.6/arch/arm/mach-pxa/mainstone.c
@@ -162,7 +162,7 @@ static void mainstone_irq_handler(unsign
 		GEDR(0) = GPIO_bit(0);  /* clear useless edge notification */
 		if (likely(pending)) {
 			irq = MAINSTONE_IRQ(0) + __ffs(pending);
-			desc = irq_desc(irq);
+			desc = irq_desc + irq;
 			desc_handle_irq(irq, desc);
 		}
 		pending = MST_INTSETCLR & mainstone_irq_enabled;
Index: linux-2.6/arch/arm/mach-pxa/pcm990-baseboard.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-pxa/pcm990-baseboard.c
+++ linux-2.6/arch/arm/mach-pxa/pcm990-baseboard.c
@@ -262,7 +262,7 @@ static void pcm990_irq_handler(unsigned
 					GPIO_bit(PCM990_CTRL_INT_IRQ_GPIO);
 		if (likely(pending)) {
 			irq = PCM027_IRQ(0) + __ffs(pending);
-			desc = irq_desc(irq);
+			desc = irq_desc + irq;
 			desc_handle_irq(irq, desc);
 		}
 		pending = (~PCM990_INTSETCLR) & pcm990_irq_enabled;
Index: linux-2.6/arch/arm/mach-s3c2410/bast-irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-s3c2410/bast-irq.c
+++ linux-2.6/arch/arm/mach-s3c2410/bast-irq.c
@@ -87,7 +87,7 @@ bast_pc104_mask(unsigned int irqno)
 static void
 bast_pc104_maskack(unsigned int irqno)
 {
-	struct irq_desc *desc = irq_desc(IRQ_ISA);
+	struct irq_desc *desc = irq_desc + IRQ_ISA;
 
 	bast_pc104_mask(irqno);
 	desc->chip->ack(IRQ_ISA);
@@ -122,7 +122,7 @@ bast_irq_pc104_demux(unsigned int irq,
 	if (unlikely(stat == 0)) {
 		/* ack if we get an irq with nothing (ie, startup) */
 
-		desc = irq_desc(IRQ_ISA);
+		desc = irq_desc + IRQ_ISA;
 		desc->chip->ack(IRQ_ISA);
 	} else {
 		/* handle the IRQ */
@@ -130,7 +130,7 @@ bast_irq_pc104_demux(unsigned int irq,
 		for (i = 0; stat != 0; i++, stat >>= 1) {
 			if (stat & 1) {
 				irqno = bast_pc104_irqs[i];
-				desc = irq_desc(irqno);
+				desc = irq_desc + irqno;
 				desc_handle_irq(irqno, desc);
 			}
 		}
Index: linux-2.6/arch/arm/mach-s3c2412/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-s3c2412/irq.c
+++ linux-2.6/arch/arm/mach-s3c2412/irq.c
@@ -123,10 +123,10 @@ static void s3c2412_irq_demux_cfsdi(unsi
 	subsrc  &= ~submsk;
 
 	if (subsrc & INTBIT(IRQ_S3C2412_SDI))
-		desc_handle_irq(IRQ_S3C2412_SDI, irq_desc(IRQ_S3C2412_SDI));
+		desc_handle_irq(IRQ_S3C2412_SDI, irq_desc + IRQ_S3C2412_SDI);
 
 	if (subsrc & INTBIT(IRQ_S3C2412_CF))
-		desc_handle_irq(IRQ_S3C2412_CF, irq_desc(IRQ_S3C2412_CF));
+		desc_handle_irq(IRQ_S3C2412_CF, irq_desc + IRQ_S3C2412_CF);
 }
 
 #define INTMSK_CFSDI	(1UL << (IRQ_S3C2412_CFSDI - IRQ_EINT0))
Index: linux-2.6/arch/arm/mach-s3c2440/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-s3c2440/irq.c
+++ linux-2.6/arch/arm/mach-s3c2440/irq.c
@@ -58,11 +58,11 @@ static void s3c_irq_demux_wdtac97(unsign
 
 	if (subsrc != 0) {
 		if (subsrc & 1) {
-			mydesc = irq_desc(IRQ_S3C2440_WDT);
+			mydesc = irq_desc + IRQ_S3C2440_WDT;
 			desc_handle_irq(IRQ_S3C2440_WDT, mydesc);
 		}
 		if (subsrc & 2) {
-			mydesc = irq_desc(IRQ_S3C2440_AC97);
+			mydesc = irq_desc + IRQ_S3C2440_AC97;
 			desc_handle_irq(IRQ_S3C2440_AC97, mydesc);
 		}
 	}
Index: linux-2.6/arch/arm/mach-s3c2443/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-s3c2443/irq.c
+++ linux-2.6/arch/arm/mach-s3c2443/irq.c
@@ -57,13 +57,13 @@ static inline void s3c2443_irq_demux(uns
 	subsrc  &= (1 << len)-1;
 
 	end = len + irq;
-	mydesc = irq_desc(irq);
+	mydesc = irq_desc + irq;
 
 	for (; irq < end && subsrc; irq++) {
 		if (subsrc & 1)
 			desc_handle_irq(irq, mydesc);
 
-		mydesc = irq_desc(irq);
+		mydesc++;
 		subsrc >>= 1;
 	}
 }
Index: linux-2.6/arch/arm/mach-sa1100/h3600.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-sa1100/h3600.c
+++ linux-2.6/arch/arm/mach-sa1100/h3600.c
@@ -719,14 +719,14 @@ static void h3800_IRQ_demux(unsigned int
 		if (0) printk("%s KPIO 0x%08X\n", __func__, irq);
 		for (j = 0; j < H3800_KPIO_IRQ_COUNT; j++)
 			if (irq & kpio_irq_mask[j])
-				handle_edge_irq(H3800_KPIO_IRQ_COUNT + j, irq_desc(H3800_KPIO_IRQ_COUNT + j));
+				handle_edge_irq(H3800_KPIO_IRQ_COUNT + j, irq_desc + H3800_KPIO_IRQ_COUNT + j);
 
 		/* GPIO2 */
 		irq = H3800_ASIC2_GPIINTFLAG;
 		if (0) printk("%s GPIO 0x%08X\n", __func__, irq);
 		for (j = 0; j < H3800_GPIO_IRQ_COUNT; j++)
 			if (irq & gpio_irq_mask[j])
-				handle_edge_irq(H3800_GPIO_IRQ_COUNT + j, irq_desc(H3800_GPIO_IRQ_COUNT + j));
+				handle_edge_irq(H3800_GPIO_IRQ_COUNT + j, irq_desc + H3800_GPIO_IRQ_COUNT + j);
 	}
 
 	if (i >= MAX_ASIC_ISR_LOOPS)
@@ -822,17 +822,15 @@ static void __init h3800_init_irq(void)
 #if 0
 	for (i = 0; i < H3800_KPIO_IRQ_COUNT; i++) {
 		int irq = i + H3800_KPIO_IRQ_START;
-		struct irq_desc *desc = irq_desc(irq);
-		desc->valid    = 1;
-		desc->probe_ok = 1;
+		irq_desc[irq].valid    = 1;
+		irq_desc[irq].probe_ok = 1;
 		set_irq_chip(irq, &h3800_kpio_irqchip);
 	}
 
 	for (i = 0; i < H3800_GPIO_IRQ_COUNT; i++) {
 		int irq = i + H3800_GPIO_IRQ_START;
-		struct irq_desc *desc = irq_desc(irq);
-		desc->valid    = 1;
-		desc->probe_ok = 1;
+		irq_desc[irq].valid    = 1;
+		irq_desc[irq].probe_ok = 1;
 		set_irq_chip(irq, &h3800_gpio_irqchip);
 	}
 #endif
Index: linux-2.6/arch/arm/mach-sa1100/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-sa1100/irq.c
+++ linux-2.6/arch/arm/mach-sa1100/irq.c
@@ -122,7 +122,7 @@ sa1100_high_gpio_handler(unsigned int ir
 		GEDR = mask;
 
 		irq = IRQ_GPIO11;
-		desc = irq_desc(irq);
+		desc = irq_desc + irq;
 		mask >>= 11;
 		do {
 			if (mask & 1)
Index: linux-2.6/arch/arm/mach-sa1100/neponset.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-sa1100/neponset.c
+++ linux-2.6/arch/arm/mach-sa1100/neponset.c
@@ -67,12 +67,12 @@ neponset_irq_handler(unsigned int irq, s
 			desc->chip->ack(irq);
 
 			if (irr & IRR_ETHERNET) {
-				d = irq_desc(IRQ_NEPONSET_SMC9196);
+				d = irq_desc + IRQ_NEPONSET_SMC9196;
 				desc_handle_irq(IRQ_NEPONSET_SMC9196, d);
 			}
 
 			if (irr & IRR_USAR) {
-				d = irq_desc(IRQ_NEPONSET_USAR);
+				d = irq_desc + IRQ_NEPONSET_USAR;
 				desc_handle_irq(IRQ_NEPONSET_USAR, d);
 			}
 
@@ -80,7 +80,7 @@ neponset_irq_handler(unsigned int irq, s
 		}
 
 		if (irr & IRR_SA1111) {
-			d = irq_desc(IRQ_NEPONSET_SA1111);
+			d = irq_desc + IRQ_NEPONSET_SA1111;
 			desc_handle_irq(IRQ_NEPONSET_SA1111, d);
 		}
 	}
Index: linux-2.6/arch/arm/mach-versatile/core.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-versatile/core.c
+++ linux-2.6/arch/arm/mach-versatile/core.c
@@ -95,7 +95,7 @@ sic_handle_irq(unsigned int irq, struct
 
 		irq += IRQ_SIC_START;
 
-		desc = irq_desc(irq);
+		desc = irq_desc + irq;
 		desc_handle_irq(irq, desc);
 	} while (status);
 }
Index: linux-2.6/arch/arm/oprofile/op_model_mpcore.c
===================================================================
--- linux-2.6.orig/arch/arm/oprofile/op_model_mpcore.c
+++ linux-2.6/arch/arm/oprofile/op_model_mpcore.c
@@ -259,7 +259,7 @@ static void em_stop(void)
  */
 static void em_route_irq(int irq, unsigned int cpu)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	cpumask_t mask = cpumask_of_cpu(cpu);
 
 	spin_lock_irq(&desc->lock);
Index: linux-2.6/arch/arm/plat-mxc/gpio.c
===================================================================
--- linux-2.6.orig/arch/arm/plat-mxc/gpio.c
+++ linux-2.6/arch/arm/plat-mxc/gpio.c
@@ -105,14 +105,13 @@ static void mxc_gpio_irq_handler(struct
 
 	gpio_irq_no = port->virtual_irq_start;
 	for (; irq_stat != 0; irq_stat >>= 1, gpio_irq_no++) {
-		struct irq_desc *desc;
 
 		if ((irq_stat & 1) == 0)
 			continue;
 
-		desc = irq_desc(gpio_irq_no);
-		BUG_ON(!(desc->handle_irq));
-		desc->handle_irq(gpio_irq_no, desc);
+		BUG_ON(!(irq_desc[gpio_irq_no].handle_irq));
+		irq_desc[gpio_irq_no].handle_irq(gpio_irq_no,
+				&irq_desc[gpio_irq_no]);
 	}
 }
 
Index: linux-2.6/arch/arm/plat-omap/gpio.c
===================================================================
--- linux-2.6.orig/arch/arm/plat-omap/gpio.c
+++ linux-2.6/arch/arm/plat-omap/gpio.c
@@ -648,10 +648,8 @@ static int gpio_irq_type(unsigned irq, u
 	spin_lock_irqsave(&bank->lock, flags);
 	retval = _set_gpio_triggering(bank, get_gpio_index(gpio), type);
 	if (retval == 0) {
-		struct irq_desc *desc = irq_desc(irq);
-
-		desc->status &= ~IRQ_TYPE_SENSE_MASK;
-		desc->status |= type;
+		irq_desc[irq].status &= ~IRQ_TYPE_SENSE_MASK;
+		irq_desc[irq].status |= type;
 	}
 	spin_unlock_irqrestore(&bank->lock, flags);
 
@@ -1057,7 +1055,7 @@ static void gpio_irq_handler(unsigned in
 
 			if (!(isr & 1))
 				continue;
-			d = irq_desc(gpio_irq);
+			d = irq_desc + gpio_irq;
 
 			desc_handle_irq(gpio_irq, d);
 		}
@@ -1505,7 +1503,7 @@ static int __init _omap_gpio_init(void)
 
 		for (j = bank->virtual_irq_start;
 		     j < bank->virtual_irq_start + gpio_count; j++) {
-			lockdep_set_class(&irq_desc(j)->lock, &gpio_lock_class);
+			lockdep_set_class(&irq_desc[j].lock, &gpio_lock_class);
 			set_irq_chip_data(j, bank);
 			if (bank_is_mpuio(bank))
 				set_irq_chip(j, &mpuio_irq_chip);
@@ -1820,7 +1818,7 @@ static int dbg_gpio_show(struct seq_file
 
 /* FIXME for at least omap2, show pullup/pulldown state */
 
-			irqstat = irq_desc(irq)->status;
+			irqstat = irq_desc[irq].status;
 			if (is_in && ((bank->suspend_wakeup & mask)
 					|| irqstat & IRQ_TYPE_SENSE_MASK)) {
 				char	*trigger = NULL;
Index: linux-2.6/arch/arm/plat-orion/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/plat-orion/irq.c
+++ linux-2.6/arch/arm/plat-orion/irq.c
@@ -59,7 +59,7 @@ void __init orion_irq_init(unsigned int
 		set_irq_chip(irq, &orion_irq_chip);
 		set_irq_chip_data(irq, maskaddr);
 		set_irq_handler(irq, handle_level_irq);
-		irq_desc(irq)->status |= IRQ_LEVEL;
+		irq_desc[irq].status |= IRQ_LEVEL;
 		set_irq_flags(irq, IRQF_VALID);
 	}
 }
Index: linux-2.6/arch/arm/plat-s3c24xx/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/plat-s3c24xx/irq.c
+++ linux-2.6/arch/arm/plat-s3c24xx/irq.c
@@ -482,11 +482,11 @@ static void s3c_irq_demux_adc(unsigned i
 
 	if (subsrc != 0) {
 		if (subsrc & 1) {
-			mydesc = irq_desc(IRQ_TC);
+			mydesc = irq_desc + IRQ_TC;
 			desc_handle_irq(IRQ_TC, mydesc);
 		}
 		if (subsrc & 2) {
-			mydesc = irq_desc(IRQ_ADC);
+			mydesc = irq_desc + IRQ_ADC;
 			desc_handle_irq(IRQ_ADC, mydesc);
 		}
 	}
@@ -512,17 +512,17 @@ static void s3c_irq_demux_uart(unsigned
 	subsrc &= 7;
 
 	if (subsrc != 0) {
-		desc = irq_desc(start);
+		desc = irq_desc + start;
 
 		if (subsrc & 1)
 			desc_handle_irq(start, desc);
 
-		desc = irq_desc(start + 1);
+		desc++;
 
 		if (subsrc & 2)
 			desc_handle_irq(start+1, desc);
 
-		desc = irq_desc(start + 2);
+		desc++;
 
 		if (subsrc & 4)
 			desc_handle_irq(start+2, desc);
@@ -572,7 +572,7 @@ s3c_irq_demux_extint8(unsigned int irq,
 		eintpnd &= ~(1<<irq);
 
 		irq += (IRQ_EINT4 - 4);
-		desc_handle_irq(irq, irq_desc(irq));
+		desc_handle_irq(irq, irq_desc + irq);
 	}
 
 }
@@ -595,7 +595,7 @@ s3c_irq_demux_extint4t7(unsigned int irq
 
 		irq += (IRQ_EINT4 - 4);
 
-		desc_handle_irq(irq, irq_desc(irq));
+		desc_handle_irq(irq, irq_desc + irq);
 	}
 }
 
Index: linux-2.6/arch/arm/plat-s3c24xx/s3c244x-irq.c
===================================================================
--- linux-2.6.orig/arch/arm/plat-s3c24xx/s3c244x-irq.c
+++ linux-2.6/arch/arm/plat-s3c24xx/s3c244x-irq.c
@@ -58,11 +58,11 @@ static void s3c_irq_demux_cam(unsigned i
 
 	if (subsrc != 0) {
 		if (subsrc & 1) {
-			mydesc = irq_desc(IRQ_S3C2440_CAM_C);
+			mydesc = irq_desc + IRQ_S3C2440_CAM_C;
 			desc_handle_irq(IRQ_S3C2440_CAM_C, mydesc);
 		}
 		if (subsrc & 2) {
-			mydesc = irq_desc(IRQ_S3C2440_CAM_P);
+			mydesc = irq_desc + IRQ_S3C2440_CAM_P;
 			desc_handle_irq(IRQ_S3C2440_CAM_P, mydesc);
 		}
 	}
Index: linux-2.6/arch/avr32/mach-at32ap/extint.c
===================================================================
--- linux-2.6.orig/arch/avr32/mach-at32ap/extint.c
+++ linux-2.6/arch/avr32/mach-at32ap/extint.c
@@ -88,6 +88,7 @@ static void eic_unmask_irq(unsigned int
 static int eic_set_irq_type(unsigned int irq, unsigned int flow_type)
 {
 	struct eic *eic = get_irq_chip_data(irq);
+	struct irq_desc *desc;
 	unsigned int i = irq - eic->first_irq;
 	u32 mode, edge, level;
 	int ret = 0;
@@ -96,6 +97,8 @@ static int eic_set_irq_type(unsigned int
 	if (flow_type == IRQ_TYPE_NONE)
 		flow_type = IRQ_TYPE_LEVEL_LOW;
 
+	desc = &irq_desc[irq];
+
 	mode = eic_readl(eic, MODE);
 	edge = eic_readl(eic, EDGE);
 	level = eic_readl(eic, LEVEL);
@@ -123,10 +126,6 @@ static int eic_set_irq_type(unsigned int
 	}
 
 	if (ret == 0) {
-		struct irq_desc *desc;
-
-		desc = irq_desc(irq);
-
 		eic_writel(eic, MODE, mode);
 		eic_writel(eic, EDGE, edge);
 		eic_writel(eic, LEVEL, level);
Index: linux-2.6/arch/avr32/mach-at32ap/intc.c
===================================================================
--- linux-2.6.orig/arch/avr32/mach-at32ap/intc.c
+++ linux-2.6/arch/avr32/mach-at32ap/intc.c
@@ -69,7 +69,7 @@ asmlinkage void do_IRQ(int level, struct
 	irq_enter();
 
 	irq = intc_readl(&intc0, INTCAUSE0 - 4 * level);
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	desc->handle_irq(irq, desc);
 
 	/*
Index: linux-2.6/arch/avr32/mach-at32ap/pio.c
===================================================================
--- linux-2.6.orig/arch/avr32/mach-at32ap/pio.c
+++ linux-2.6/arch/avr32/mach-at32ap/pio.c
@@ -262,7 +262,7 @@ static void gpio_irq_handler(unsigned ir
 			isr &= ~(1 << i);
 
 			i += gpio_irq;
-			d = irq_desc(i);
+			d = &irq_desc[i];
 
 			d->handle_irq(i, d);
 		} while (isr);
Index: linux-2.6/arch/blackfin/kernel/irqchip.c
===================================================================
--- linux-2.6.orig/arch/blackfin/kernel/irqchip.c
+++ linux-2.6/arch/blackfin/kernel/irqchip.c
@@ -77,10 +77,8 @@ int show_interrupts(struct seq_file *p,
 	unsigned long flags;
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 
@@ -91,7 +89,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
  unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 		seq_printf(p, "Err: %10lu\n", irq_err_count);
 	}
@@ -110,7 +108,7 @@ __attribute__((l1_text))
 asmlinkage void asm_do_IRQ(unsigned int irq, struct pt_regs *regs)
 {
 	struct pt_regs *old_regs;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	unsigned short pending, other_ints;
 
 	old_regs = set_irq_regs(regs);
@@ -146,7 +144,7 @@ void __init init_IRQ(void)
 	int irq;
 
 	spin_lock_init(&irq_controller_lock);
-	for (irq = 0, desc = irq_desc(0); irq < NR_IRQS; irq++, desc = irq_desc(irq)) {
+	for (irq = 0, desc = irq_desc; irq < NR_IRQS; irq++, desc++) {
 		*desc = bad_irq_desc;
 	}
 
Index: linux-2.6/arch/blackfin/kernel/traps.c
===================================================================
--- linux-2.6.orig/arch/blackfin/kernel/traps.c
+++ linux-2.6/arch/blackfin/kernel/traps.c
@@ -1003,10 +1003,8 @@ void show_regs(struct pt_regs *fp)
 	/* if no interrupts are going off, don't print this out */
 	if (fp->ipend & ~0x3F) {
 		for (i = 0; i < (NR_IRQS - 1); i++) {
-			struct irq_desc *desc = irq_desc(i);
-
-			spin_lock_irqsave(&desc->lock, flags);
-			action = desc->action;
+			spin_lock_irqsave(&irq_desc[i].lock, flags);
+			action = irq_desc[i].action;
 			if (!action)
 				goto unlock;
 
@@ -1018,7 +1016,7 @@ void show_regs(struct pt_regs *fp)
 			}
 			printk("\n");
 unlock:
-			spin_unlock_irqrestore(&desc->lock, flags);
+			spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 		}
 	}
 
Index: linux-2.6/arch/blackfin/mach-bf537/boards/cm_bf537.c
===================================================================
--- linux-2.6.orig/arch/blackfin/mach-bf537/boards/cm_bf537.c
+++ linux-2.6/arch/blackfin/mach-bf537/boards/cm_bf537.c
@@ -575,7 +575,7 @@ static int __init cm_bf537_init(void)
 #endif
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
+	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
Index: linux-2.6/arch/blackfin/mach-bf537/boards/stamp.c
===================================================================
--- linux-2.6.orig/arch/blackfin/mach-bf537/boards/stamp.c
+++ linux-2.6/arch/blackfin/mach-bf537/boards/stamp.c
@@ -1026,7 +1026,7 @@ static int __init stamp_init(void)
 	spi_register_board_info(bfin_spi_board_info, ARRAY_SIZE(bfin_spi_board_info));
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
+	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
 #endif
 
 	return 0;
Index: linux-2.6/arch/blackfin/mach-bf561/boards/cm_bf561.c
===================================================================
--- linux-2.6.orig/arch/blackfin/mach-bf561/boards/cm_bf561.c
+++ linux-2.6/arch/blackfin/mach-bf561/boards/cm_bf561.c
@@ -411,7 +411,7 @@ static int __init cm_bf561_init(void)
 #endif
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
+	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
Index: linux-2.6/arch/blackfin/mach-common/ints-priority.c
===================================================================
--- linux-2.6.orig/arch/blackfin/mach-common/ints-priority.c
+++ linux-2.6/arch/blackfin/mach-common/ints-priority.c
@@ -315,7 +315,7 @@ static void bfin_demux_error_irq(unsigne
 
 	if (irq) {
 		if (error_int_mask & (1L << (irq - IRQ_PPI_ERROR))) {
-			struct irq_desc *desc = irq_desc(irq);
+			struct irq_desc *desc = irq_desc + irq;
 			desc->handle_irq(irq, desc);
 		} else {
 
@@ -561,7 +561,7 @@ static void bfin_demux_gpio_irq(unsigned
 
 			while (mask) {
 				if (mask & 1) {
-					desc = irq_desc(irq);
+					desc = irq_desc + irq;
 					desc->handle_irq(irq, desc);
 				}
 				irq++;
@@ -576,7 +576,7 @@ static void bfin_demux_gpio_irq(unsigned
 
 			do {
 				if (mask & 1) {
-					desc = irq_desc(irq);
+					desc = irq_desc + irq;
 					desc->handle_irq(irq, desc);
 				}
 				irq++;
@@ -928,7 +928,7 @@ static void bfin_demux_gpio_irq(unsigned
 	while (request) {
 		if (request & 1) {
 			irq = pint2irq_lut[pint_val] + SYS_IRQS;
-			desc = irq_desc(irq);
+			desc = irq_desc + irq;
 			desc->handle_irq(irq, desc);
 		}
 		pint_val++;
Index: linux-2.6/arch/cris/arch-v10/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/cris/arch-v10/kernel/irq.c
+++ linux-2.6/arch/cris/arch-v10/kernel/irq.c
@@ -221,7 +221,7 @@ init_IRQ(void)
 
 	/* Initialize IRQ handler descriptors. */
 	for(i = 2; i < NR_IRQS; i++) {
-		irq_desc(i)->chip = &crisv10_irq_type;
+		irq_desc[i].chip = &crisv10_irq_type;
 		set_int_vector(i, interrupt[i]);
 	}
 
Index: linux-2.6/arch/cris/arch-v32/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/cris/arch-v32/kernel/irq.c
+++ linux-2.6/arch/cris/arch-v32/kernel/irq.c
@@ -475,15 +475,15 @@ init_IRQ(void)
 
 	/* Point all IRQ's to bad handlers. */
 	for (i = FIRST_IRQ, j = 0; j < NR_IRQS; i++, j++) {
-		irq_desc(j)->chip = &crisv32_irq_type;
+		irq_desc[j].chip = &crisv32_irq_type;
 		set_exception_vector(i, interrupt[j]);
 	}
 
         /* Mark Timer and IPI IRQs as CPU local */
 	irq_allocations[TIMER0_INTR_VECT - FIRST_IRQ].cpu = CPU_FIXED;
-	irq_desc(TIMER0_INTR_VECT)->status |= IRQ_PER_CPU;
+	irq_desc[TIMER0_INTR_VECT].status |= IRQ_PER_CPU;
 	irq_allocations[IPI_INTR_VECT - FIRST_IRQ].cpu = CPU_FIXED;
-	irq_desc(IPI_INTR_VECT)->status |= IRQ_PER_CPU;
+	irq_desc[IPI_INTR_VECT].status |= IRQ_PER_CPU;
 
 	set_exception_vector(0x00, nmi_interrupt);
 	set_exception_vector(0x30, multiple_interrupt);
Index: linux-2.6/arch/cris/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/cris/kernel/irq.c
+++ linux-2.6/arch/cris/kernel/irq.c
@@ -57,10 +57,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ",i);
@@ -70,7 +68,7 @@ int show_interrupts(struct seq_file *p,
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
-		seq_printf(p, " %14s", desc->chip->typename);
+		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -78,7 +76,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	}
 	return 0;
 }
@@ -101,7 +99,7 @@ asmlinkage void do_IRQ(int irq, struct p
 		printk("do_IRQ: stack overflow: %lX\n", sp);
 		show_stack(NULL, (unsigned long *)sp);
 	}
-	__do_IRQ(irq, irq_desc(irq));
+	__do_IRQ(irq, &irq_desc[irq]);
         irq_exit();
 	set_irq_regs(old_regs);
 }
Index: linux-2.6/arch/frv/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/frv/kernel/irq.c
+++ linux-2.6/arch/frv/kernel/irq.c
@@ -69,15 +69,13 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (action) {
 			seq_printf(p, "%3d: ", i);
 			for_each_present_cpu(cpu)
 				seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
-			seq_printf(p, " %10s", desc->chip->name ? : "-");
+			seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
 			seq_printf(p, "  %s", action->name);
 			for (action = action->next;
 			     action;
@@ -87,7 +85,7 @@ int show_interrupts(struct seq_file *p,
 			seq_putc(p, '\n');
 		}
 
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 		seq_printf(p, "Err: %10u\n", atomic_read(&irq_err_count));
 	}
Index: linux-2.6/arch/h8300/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/h8300/kernel/irq.c
+++ linux-2.6/arch/h8300/kernel/irq.c
@@ -166,19 +166,17 @@ void __init init_IRQ(void)
 	setup_vector();
 
 	for (c = 0; c < NR_IRQS; c++) {
-		struct irq_desc *desc = irq_desc(c);
-
-		desc->status = IRQ_DISABLED;
-		desc->action = NULL;
-		desc->depth = 1;
-		desc->chip = &h8300irq_chip;
+		irq_desc[c].status = IRQ_DISABLED;
+		irq_desc[c].action = NULL;
+		irq_desc[c].depth = 1;
+		irq_desc[c].chip = &h8300irq_chip;
 	}
 }
 
 asmlinkage void do_IRQ(int irq)
 {
 	irq_enter();
-	__do_IRQ(irq, irq_desc(irq));
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 }
 
@@ -193,23 +191,21 @@ int show_interrupts(struct seq_file *p,
 		seq_puts(p, "           CPU0");
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 		seq_printf(p, "%3d: ",i);
 		seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
-		seq_printf(p, " %14s", desc->chip->name);
-		seq_printf(p, "-%-8s", desc->name);
+		seq_printf(p, " %14s", irq_desc[i].chip->name);
+		seq_printf(p, "-%-8s", irq_desc[i].name);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
 		seq_putc(p, '\n');
 unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	}
 	return 0;
 }
Index: linux-2.6/arch/ia64/hp/sim/hpsim_irq.c
===================================================================
--- linux-2.6.orig/arch/ia64/hp/sim/hpsim_irq.c
+++ linux-2.6/arch/ia64/hp/sim/hpsim_irq.c
@@ -44,7 +44,7 @@ hpsim_irq_init (void)
 	int i;
 
 	for (i = 0; i < NR_IRQS; ++i) {
-		idesc = irq_desc(i);
+		idesc = irq_desc + i;
 		if (idesc->chip == &no_irq_type)
 			idesc->chip = &irq_type_hp_sim;
 	}
Index: linux-2.6/arch/ia64/include/asm/hw_irq.h
===================================================================
--- linux-2.6.orig/arch/ia64/include/asm/hw_irq.h
+++ linux-2.6/arch/ia64/include/asm/hw_irq.h
@@ -146,7 +146,7 @@ static inline void ia64_native_resend_ir
  * Default implementations for the irq-descriptor API:
  */
 
-extern irq_desc_t irq_descX[NR_IRQS];
+extern irq_desc_t irq_desc[NR_IRQS];
 
 #ifndef CONFIG_IA64_GENERIC
 static inline ia64_vector __ia64_irq_to_vector(int irq)
Index: linux-2.6/arch/ia64/kernel/iosapic.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/iosapic.c
+++ linux-2.6/arch/ia64/kernel/iosapic.c
@@ -398,7 +398,7 @@ iosapic_end_level_irq (unsigned int irq)
 	int do_unmask_irq = 0;
 
 	irq_complete_move(irq);
-	if (unlikely(irq_desc(irq)->status & IRQ_MOVE_PENDING)) {
+	if (unlikely(irq_desc[irq].status & IRQ_MOVE_PENDING)) {
 		do_unmask_irq = 1;
 		mask_irq(irq);
 	}
@@ -407,7 +407,7 @@ iosapic_end_level_irq (unsigned int irq)
 		iosapic_eoi(rte->iosapic->addr, vec);
 
 	if (unlikely(do_unmask_irq)) {
-		move_masked_irq(irq, irq_desc(irq));
+		move_masked_irq(irq, &irq_desc[irq]);
 		unmask_irq(irq);
 	}
 }
@@ -449,7 +449,7 @@ iosapic_startup_edge_irq (unsigned int i
 static void
 iosapic_ack_edge_irq (unsigned int irq)
 {
-	irq_desc_t *idesc = irq_desc(irq);
+	irq_desc_t *idesc = irq_desc + irq;
 
 	irq_complete_move(irq);
 	move_native_irq(irq, idesc);
@@ -646,7 +646,7 @@ register_intr (unsigned int gsi, int irq
 
 	irq_type = iosapic_get_irq_chip(trigger);
 
-	idesc = irq_desc(irq);
+	idesc = irq_desc + irq;
 	if (irq_type != NULL && idesc->chip != irq_type) {
 		if (idesc->chip != &no_irq_type)
 			printk(KERN_WARNING
@@ -765,7 +765,6 @@ iosapic_register_intr (unsigned int gsi,
 	struct iosapic_rte_info *rte;
 	u32 low32;
 	unsigned char dmode;
-	struct irq_desc *desc;
 
 	/*
 	 * If this GSI has already been registered (i.e., it's a
@@ -793,14 +792,12 @@ iosapic_register_intr (unsigned int gsi,
 			goto unlock_iosapic_lock;
 	}
 
-	desc = irq_desc(irq);
-
-	spin_lock(&desc->lock);
+	spin_lock(&irq_desc[irq].lock);
 	dest = get_target_cpu(gsi, irq);
 	dmode = choose_dmode();
 	err = register_intr(gsi, irq, dmode, polarity, trigger);
 	if (err < 0) {
-		spin_unlock(&desc->lock);
+		spin_unlock(&irq_desc[irq].lock);
 		irq = err;
 		goto unlock_iosapic_lock;
 	}
@@ -819,7 +816,7 @@ iosapic_register_intr (unsigned int gsi,
 	       (polarity == IOSAPIC_POL_HIGH ? "high" : "low"),
 	       cpu_logical_id(dest), dest, irq_to_vector(irq));
 
-	spin_unlock(&desc->lock);
+	spin_unlock(&irq_desc[irq].lock);
  unlock_iosapic_lock:
 	spin_unlock_irqrestore(&iosapic_lock, flags);
 	return irq;
@@ -860,7 +857,7 @@ iosapic_unregister_intr (unsigned int gs
 	if (--rte->refcnt > 0)
 		goto out;
 
-	idesc = irq_desc(irq);
+	idesc = irq_desc + irq;
 	rte->refcnt = NO_REF_RTE;
 
 	/* Mask the interrupt */
Index: linux-2.6/arch/ia64/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/irq.c
+++ linux-2.6/arch/ia64/kernel/irq.c
@@ -71,10 +71,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ",i);
@@ -85,7 +83,7 @@ int show_interrupts(struct seq_file *p,
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 		}
 #endif
-		seq_printf(p, " %14s", desc->chip->name);
+		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -93,7 +91,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS)
 		seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
 	return 0;
@@ -109,7 +107,7 @@ void set_irq_affinity_info (unsigned int
 	cpu_set(cpu_logical_id(hwid), mask);
 
 	if (irq < NR_IRQS) {
-		irq_desc(irq)->affinity = mask;
+		irq_desc[irq].affinity = mask;
 		irq_redir[irq] = (char) (redir & 0xff);
 	}
 }
@@ -140,7 +138,7 @@ static void migrate_irqs(void)
 	int 		irq, new_cpu;
 
 	for (irq=0; irq < NR_IRQS; irq++) {
-		desc = irq_desc(irq);
+		desc = irq_desc + irq;
 
 		if (desc->status == IRQ_DISABLED)
 			continue;
@@ -154,7 +152,7 @@ static void migrate_irqs(void)
 		if (desc->status == IRQ_PER_CPU)
 			continue;
 
-		cpus_and(mask, desc->affinity, cpu_online_map);
+		cpus_and(mask, irq_desc[irq].affinity, cpu_online_map);
 		if (any_online_cpu(mask) == NR_CPUS) {
 			/*
 			 * Save it for phase 2 processing
Index: linux-2.6/arch/ia64/kernel/irq_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/irq_ia64.c
+++ linux-2.6/arch/ia64/kernel/irq_ia64.c
@@ -343,7 +343,7 @@ static irqreturn_t smp_irq_move_cleanup_
 		if (irq < 0)
 			continue;
 
-		desc = irq_desc(irq);
+		desc = irq_desc + irq;
 		cfg = irq_cfg + irq;
 		spin_lock(&desc->lock);
 		if (!cfg->move_cleanup_count)
@@ -495,9 +495,9 @@ ia64_handle_irq (ia64_vector vector, str
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
 		if (unlikely(IS_LOCAL_TLB_FLUSH(vector))) {
 			smp_local_flush_tlb();
-			kstat_irqs_this_cpu(irq_desc(vector))++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		} else if (unlikely(IS_RESCHEDULE(vector)))
-			kstat_irqs_this_cpu(irq_desc(vector))++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		else {
 			int irq = local_vector_to_irq(vector);
 
@@ -553,9 +553,9 @@ void ia64_process_pending_intr(void)
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
 		if (unlikely(IS_LOCAL_TLB_FLUSH(vector))) {
 			smp_local_flush_tlb();
-			kstat_irqs_this_cpu(irq_desc(vector))++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		} else if (unlikely(IS_RESCHEDULE(vector)))
-			kstat_irqs_this_cpu(irq_desc(vector))++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		else {
 			struct pt_regs *old_regs = set_irq_regs(NULL);
 			int irq = local_vector_to_irq(vector);
@@ -629,7 +629,7 @@ ia64_native_register_percpu_irq (ia64_ve
 
 	irq = vec;
 	BUG_ON(bind_irq_vector(irq, vec, CPU_MASK_ALL));
-	desc = irq_desc(irq);
+	desc = irq_desc + irq;
 	desc->status |= IRQ_PER_CPU;
 	desc->chip = &irq_type_ia64_lsapic;
 	if (action)
Index: linux-2.6/arch/ia64/kernel/mca.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/mca.c
+++ linux-2.6/arch/ia64/kernel/mca.c
@@ -2101,7 +2101,7 @@ ia64_mca_late_init(void)
 			irq = local_vector_to_irq(cpe_vector);
 			if (irq > 0) {
 				cpe_poll_enabled = 0;
-				desc = irq_desc(irq);
+				desc = irq_desc + irq;
 				desc->status |= IRQ_PER_CPU;
 				setup_irq(irq, &mca_cpe_irqaction);
 				ia64_cpe_irq = irq;
Index: linux-2.6/arch/ia64/kernel/msi_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/msi_ia64.c
+++ linux-2.6/arch/ia64/kernel/msi_ia64.c
@@ -73,7 +73,7 @@ static void ia64_set_msi_irq_affinity(un
 	msg.data = data;
 
 	write_msi_msg(irq, &msg);
-	irq_desc(irq)->affinity = cpumask_of_cpu(cpu);
+	irq_desc[irq].affinity = cpumask_of_cpu(cpu);
 }
 #endif /* CONFIG_SMP */
 
@@ -120,7 +120,7 @@ void ia64_teardown_msi_irq(unsigned int
 static void ia64_ack_msi_irq(unsigned int irq)
 {
 	irq_complete_move(irq);
-	move_native_irq(irq, irq_desc(irq));
+	move_native_irq(irq, &irq_desc[irq]);
 	ia64_eoi();
 }
 
Index: linux-2.6/arch/ia64/kernel/smpboot.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/smpboot.c
+++ linux-2.6/arch/ia64/kernel/smpboot.c
@@ -701,7 +701,7 @@ int migrate_platform_irqs(unsigned int c
 			new_cpei_cpu = any_online_cpu(cpu_online_map);
 			mask = cpumask_of_cpu(new_cpei_cpu);
 			set_cpei_target_cpu(new_cpei_cpu);
-			desc = irq_desc(ia64_cpe_irq);
+			desc = irq_desc + ia64_cpe_irq;
 			/*
 			 * Switch for now, immediately, we need to do fake intr
 			 * as other interrupts, but need to study CPEI behaviour with
Index: linux-2.6/arch/ia64/sn/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/irq.c
+++ linux-2.6/arch/ia64/sn/kernel/irq.c
@@ -110,7 +110,7 @@ static void sn_ack_irq(unsigned int irq)
 	HUB_S((u64*)LOCAL_MMR_ADDR(SH_EVENT_OCCURRED_ALIAS), mask);
 	__set_bit(irq, (volatile void *)pda->sn_in_service_ivecs);
 
-	move_native_irq(irq, irq_desc(irq));
+	move_native_irq(irq, &irq_desc[irq]);
 }
 
 static void sn_end_irq(unsigned int irq)
@@ -293,15 +293,14 @@ unsigned int sn_local_vector_to_irq(u8 v
 void sn_irq_init(void)
 {
 	int i;
-	irq_desc_t *desc;
+	irq_desc_t *base_desc = irq_desc;
 
 	ia64_first_device_vector = IA64_SN2_FIRST_DEVICE_VECTOR;
 	ia64_last_device_vector = IA64_SN2_LAST_DEVICE_VECTOR;
 
 	for (i = 0; i < NR_IRQS; i++) {
-		desc = irq_desc(i);
-		if (desc->chip == &no_irq_type) {
-			desc->chip = &irq_type_sn;
+		if (base_desc[i].chip == &no_irq_type) {
+			base_desc[i].chip = &irq_type_sn;
 		}
 	}
 }
@@ -430,7 +429,7 @@ sn_call_force_intr_provider(struct sn_ir
 	pci_provider = sn_pci_provider[sn_irq_info->irq_bridge_type];
 
 	/* Don't force an interrupt if the irq has been disabled */
-	if (!(irq_desc(sn_irq_info->irq_irq)->status & IRQ_DISABLED) &&
+	if (!(irq_desc[sn_irq_info->irq_irq].status & IRQ_DISABLED) &&
 	    pci_provider && pci_provider->force_interrupt)
 		(*pci_provider->force_interrupt)(sn_irq_info);
 }
Index: linux-2.6/arch/ia64/sn/kernel/msi_sn.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/msi_sn.c
+++ linux-2.6/arch/ia64/sn/kernel/msi_sn.c
@@ -204,13 +204,13 @@ static void sn_set_msi_irq_affinity(unsi
 	msg.address_lo = (u32)(bus_addr & 0x00000000ffffffff);
 
 	write_msi_msg(irq, &msg);
-	irq_desc(irq)->affinity = cpu_mask;
+	irq_desc[irq].affinity = cpu_mask;
 }
 #endif /* CONFIG_SMP */
 
 static void sn_ack_msi_irq(unsigned int irq)
 {
-	move_native_irq(irq, irq_desc(irq));
+	move_native_irq(irq, &irq_desc[irq]);
 	ia64_eoi();
 }
 
Index: linux-2.6/arch/m32r/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m32r/kernel/irq.c
+++ linux-2.6/arch/m32r/kernel/irq.c
@@ -43,10 +43,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ",i);
@@ -56,7 +54,7 @@ int show_interrupts(struct seq_file *p,
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
-		seq_printf(p, " %14s", desc->chip->typename);
+		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -64,7 +62,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 		seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
 		seq_printf(p, "MIS: %10u\n", atomic_read(&irq_mis_count));
@@ -86,7 +84,7 @@ asmlinkage unsigned int do_IRQ(int irq,
 #ifdef CONFIG_DEBUG_STACKOVERFLOW
 	/* FIXME M32R */
 #endif
-	__do_IRQ(irq, irq_desc(irq));
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 	set_irq_regs(old_regs);
 
Index: linux-2.6/arch/m32r/platforms/m32104ut/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/m32104ut/setup.c
+++ linux-2.6/arch/m32r/platforms/m32104ut/setup.c
@@ -77,7 +77,6 @@ static struct hw_interrupt_type m32104ut
 void __init init_IRQ(void)
 {
 	static int once = 0;
-	struct irq_desc *desc;
 
 	if (once)
 		return;
@@ -86,40 +85,36 @@ void __init init_IRQ(void)
 
 #if defined(CONFIG_SMC91X)
 	/* INT#0: LAN controller on M32104UT-LAN (SMC91C111)*/
-	desc = irq_desc(M32R_IRQ_INT0);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32104ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT0].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT0].chip = &m32104ut_irq_type;
+	irq_desc[M32R_IRQ_INT0].action = 0;
+	irq_desc[M32R_IRQ_INT0].depth = 1;
 	icu_data[M32R_IRQ_INT0].icucr = M32R_ICUCR_IEN | M32R_ICUCR_ISMOD11; /* "H" level sense */
 	disable_m32104ut_irq(M32R_IRQ_INT0);
 #endif  /* CONFIG_SMC91X */
 
 	/* MFT2 : system timer */
-	desc = irq_desc(M32R_IRQ_MFT2);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32104ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &m32104ut_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_m32104ut_irq(M32R_IRQ_MFT2);
 
 #ifdef CONFIG_SERIAL_M32R_SIO
 	/* SIO0_R : uart receive data */
-	desc = irq_desc(M32R_IRQ_SIO0_R);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32104ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &m32104ut_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = M32R_ICUCR_IEN;
 	disable_m32104ut_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0_S : uart send data */
-	desc = irq_desc(M32R_IRQ_SIO0_S);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32104ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &m32104ut_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = M32R_ICUCR_IEN;
 	disable_m32104ut_irq(M32R_IRQ_SIO0_S);
 #endif /* CONFIG_SERIAL_M32R_SIO */
Index: linux-2.6/arch/m32r/platforms/m32700ut/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/m32700ut/setup.c
+++ linux-2.6/arch/m32r/platforms/m32700ut/setup.c
@@ -297,117 +297,103 @@ static struct hw_interrupt_type m32700ut
 
 void __init init_IRQ(void)
 {
-	struct irq_desc *desc;
-
 #if defined(CONFIG_SMC91X)
 	/* INT#0: LAN controller on M32700UT-LAN (SMC91C111)*/
-	desc = irq_desc(M32700UT_LAN_IRQ_LAN);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32700UT_LAN_IRQ_LAN].status = IRQ_DISABLED;
+	irq_desc[M32700UT_LAN_IRQ_LAN].chip = &m32700ut_lanpld_irq_type;
+	irq_desc[M32700UT_LAN_IRQ_LAN].action = 0;
+	irq_desc[M32700UT_LAN_IRQ_LAN].depth = 1;	/* disable nested irq */
 	lanpld_icu_data[irq2lanpldirq(M32700UT_LAN_IRQ_LAN)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD02;	/* "H" edge sense */
 	disable_m32700ut_lanpld_irq(M32700UT_LAN_IRQ_LAN);
 #endif  /* CONFIG_SMC91X */
 
 	/* MFT2 : system timer */
-	desc = irq_desc(M32R_IRQ_MFT2);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_m32700ut_irq(M32R_IRQ_MFT2);
 
 	/* SIO0 : receive */
-	desc = irq_desc(M32R_IRQ_SIO0_R);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_m32700ut_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0 : send */
-	desc = irq_desc(M32R_IRQ_SIO0_S);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_m32700ut_irq(M32R_IRQ_SIO0_S);
 
 	/* SIO1 : receive */
-	desc = irq_desc(M32R_IRQ_SIO1_R);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = 0;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_m32700ut_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1 : send */
-	desc = irq_desc(M32R_IRQ_SIO1_S);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = 0;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_m32700ut_irq(M32R_IRQ_SIO1_S);
 
 	/* DMA1 : */
-	desc = irq_desc(M32R_IRQ_DMA1)
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_DMA1].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_DMA1].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_DMA1].action = 0;
+	irq_desc[M32R_IRQ_DMA1].depth = 1;
 	icu_data[M32R_IRQ_DMA1].icucr = 0;
 	disable_m32700ut_irq(M32R_IRQ_DMA1);
 
 #ifdef CONFIG_SERIAL_M32R_PLDSIO
 	/* INT#1: SIO0 Receive on PLD */
-	desc = irq_desc(PLD_IRQ_SIO0_RCV);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_SIO0_RCV].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_SIO0_RCV].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_SIO0_RCV].action = 0;
+	irq_desc[PLD_IRQ_SIO0_RCV].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_SIO0_RCV)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD03;
 	disable_m32700ut_pld_irq(PLD_IRQ_SIO0_RCV);
 
 	/* INT#1: SIO0 Send on PLD */
-	desc = irq_desc(PLD_IRQ_SIO0_SND);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_SIO0_SND].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_SIO0_SND].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_SIO0_SND].action = 0;
+	irq_desc[PLD_IRQ_SIO0_SND].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_SIO0_SND)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD03;
 	disable_m32700ut_pld_irq(PLD_IRQ_SIO0_SND);
 #endif  /* CONFIG_SERIAL_M32R_PLDSIO */
 
 	/* INT#1: CFC IREQ on PLD */
-	desc = irq_desc(PLD_IRQ_CFIREQ);
-	desc->status = IRQ_DISABLED;
-	desc->hip = &m32700ut_pld_irq_type;
-	desc->ction = 0;
-	desc->epth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_CFIREQ].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFIREQ].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_CFIREQ].action = 0;
+	irq_desc[PLD_IRQ_CFIREQ].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_CFIREQ)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD01;	/* 'L' level sense */
 	disable_m32700ut_pld_irq(PLD_IRQ_CFIREQ);
 
 	/* INT#1: CFC Insert on PLD */
-	desc = irq_desc(PLD_IRQ_CFC_INSERT);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_CFC_INSERT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_INSERT].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_CFC_INSERT].action = 0;
+	irq_desc[PLD_IRQ_CFC_INSERT].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_CFC_INSERT)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD00;	/* 'L' edge sense */
 	disable_m32700ut_pld_irq(PLD_IRQ_CFC_INSERT);
 
 	/* INT#1: CFC Eject on PLD */
-	desc = irq_desc(PLD_IRQ_CFC_EJECT);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_CFC_EJECT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_EJECT].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_CFC_EJECT].action = 0;
+	irq_desc[PLD_IRQ_CFC_EJECT].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_CFC_EJECT)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD02;	/* 'H' edge sense */
 	disable_m32700ut_pld_irq(PLD_IRQ_CFC_EJECT);
 
@@ -427,13 +413,13 @@ void __init init_IRQ(void)
 
 #if defined(CONFIG_USB)
 	outw(USBCR_OTGS, USBCR); 	/* USBCR: non-OTG */
-	desc = irq_desc(M32700UT_LCD_IRQ_USB_INT1);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_lcdpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
-	lcdpld_icu_data[irq2lcdpldirq(M32700UT_LCD_IRQ_USB_INT1)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD01;	/* "L" level sense */
-	disable_m32700ut_lcdpld_irq(M32700UT_LCD_IRQ_USB_INT1);
+
+    irq_desc[M32700UT_LCD_IRQ_USB_INT1].status = IRQ_DISABLED;
+    irq_desc[M32700UT_LCD_IRQ_USB_INT1].chip = &m32700ut_lcdpld_irq_type;
+    irq_desc[M32700UT_LCD_IRQ_USB_INT1].action = 0;
+    irq_desc[M32700UT_LCD_IRQ_USB_INT1].depth = 1;
+    lcdpld_icu_data[irq2lcdpldirq(M32700UT_LCD_IRQ_USB_INT1)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD01;	/* "L" level sense */
+    disable_m32700ut_lcdpld_irq(M32700UT_LCD_IRQ_USB_INT1);
 #endif
 	/*
 	 * INT2# is used for BAT, USB, AUDIO
@@ -446,11 +432,10 @@ void __init init_IRQ(void)
 	/*
 	 * INT3# is used for AR
 	 */
-	desc = irq_desc(M32R_IRQ_INT3);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT3].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT3].chip = &m32700ut_irq_type;
+	irq_desc[M32R_IRQ_INT3].action = 0;
+	irq_desc[M32R_IRQ_INT3].depth = 1;
 	icu_data[M32R_IRQ_INT3].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_m32700ut_irq(M32R_IRQ_INT3);
 #endif	/* CONFIG_VIDEO_M32R_AR */
Index: linux-2.6/arch/m32r/platforms/mappi/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/mappi/setup.c
+++ linux-2.6/arch/m32r/platforms/mappi/setup.c
@@ -45,7 +45,7 @@ static void mask_and_ack_mappi(unsigned
 
 static void end_mappi_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		enable_mappi_irq(irq);
 }
 
@@ -77,8 +77,6 @@ static struct hw_interrupt_type mappi_ir
 void __init init_IRQ(void)
 {
 	static int once = 0;
-	struct irq_desc *desc;
-	unsigned irq;
 
 	if (once)
 		return;
@@ -87,86 +85,70 @@ void __init init_IRQ(void)
 
 #ifdef CONFIG_NE2000
 	/* INT0 : LAN controller (RTL8019AS) */
-	irq = M32R_IRQ_INT0;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT0].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT0].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_INT0].action = NULL;
+	irq_desc[M32R_IRQ_INT0].depth = 1;
 	icu_data[M32R_IRQ_INT0].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD11;
 	disable_mappi_irq(M32R_IRQ_INT0);
 #endif /* CONFIG_M32R_NE2000 */
 
 	/* MFT2 : system timer */
-	irq = M32R_IRQ_MFT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = NULL;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_mappi_irq(M32R_IRQ_MFT2);
 
 #ifdef CONFIG_SERIAL_M32R_SIO
 	/* SIO0_R : uart receive data */
-	irq = M32R_IRQ_SIO0_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = NULL;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0_S : uart send data */
-	irq = M32R_IRQ_SIO0_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = NULL;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO0_S);
 
 	/* SIO1_R : uart receive data */
-	irq = M32R_IRQ_SIO1_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = NULL;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1_S : uart send data */
-	irq = M32R_IRQ_SIO1_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = NULL;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO1_S);
 #endif /* CONFIG_SERIAL_M32R_SIO */
 
 #if defined(CONFIG_M32R_PCC)
 	/* INT1 : pccard0 interrupt */
-	irq = M32R_IRQ_INT1;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT1].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT1].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_INT1].action = NULL;
+	irq_desc[M32R_IRQ_INT1].depth = 1;
 	icu_data[M32R_IRQ_INT1].icucr = M32R_ICUCR_IEN | M32R_ICUCR_ISMOD00;
 	disable_mappi_irq(M32R_IRQ_INT1);
 
 	/* INT2 : pccard1 interrupt */
-	irq = M32R_IRQ_INT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = NULL;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT2].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_INT2].action = NULL;
+	irq_desc[M32R_IRQ_INT2].depth = 1;
 	icu_data[M32R_IRQ_INT2].icucr = M32R_ICUCR_IEN | M32R_ICUCR_ISMOD00;
 	disable_mappi_irq(M32R_IRQ_INT2);
 #endif /* CONFIG_M32RPCC */
Index: linux-2.6/arch/m32r/platforms/mappi2/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/mappi2/setup.c
+++ linux-2.6/arch/m32r/platforms/mappi2/setup.c
@@ -83,112 +83,89 @@ static struct hw_interrupt_type mappi2_i
 
 void __init init_IRQ(void)
 {
-	struct irq_desc *desc;
-	unsigned int i;
 #if defined(CONFIG_SMC91X)
 	/* INT0 : LAN controller (SMC91111) */
-
-	irq = M32R_IRQ_INT0;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT0].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT0].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_INT0].action = 0;
+	irq_desc[M32R_IRQ_INT0].depth = 1;
 	icu_data[M32R_IRQ_INT0].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_mappi2_irq(M32R_IRQ_INT0);
 #endif  /* CONFIG_SMC91X */
 
 	/* MFT2 : system timer */
-	irq = M32R_IRQ_MFT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_mappi2_irq(M32R_IRQ_MFT2);
 
 #ifdef CONFIG_SERIAL_M32R_SIO
 	/* SIO0_R : uart receive data */
-	irq = M32R_IRQ_SIO0_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_mappi2_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0_S : uart send data */
-	irq = M32R_IRQ_SIO0_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_mappi2_irq(M32R_IRQ_SIO0_S);
 	/* SIO1_R : uart receive data */
-	irq = M32R_IRQ_SIO1_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = 0;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_mappi2_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1_S : uart send data */
-	irq = M32R_IRQ_SIO1_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = 0;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_mappi2_irq(M32R_IRQ_SIO1_S);
 #endif  /* CONFIG_M32R_USE_DBG_CONSOLE */
 
 #if defined(CONFIG_USB)
 	/* INT1 : USB Host controller interrupt */
-	irq = M32R_IRQ_INT1;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT1].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT1].chip = &mappi2_irq_type;
+	irq_desc[M32R_IRQ_INT1].action = 0;
+	irq_desc[M32R_IRQ_INT1].depth = 1;
 	icu_data[M32R_IRQ_INT1].icucr = M32R_ICUCR_ISMOD01;
 	disable_mappi2_irq(M32R_IRQ_INT1);
 #endif /* CONFIG_USB */
 
 	/* ICUCR40: CFC IREQ */
-	irq_desc PLD_IRQ_CFIREQ;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_CFIREQ].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFIREQ].chip = &mappi2_irq_type;
+	irq_desc[PLD_IRQ_CFIREQ].action = 0;
+	irq_desc[PLD_IRQ_CFIREQ].depth = 1;	/* disable nested irq */
 	icu_data[PLD_IRQ_CFIREQ].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD01;
 	disable_mappi2_irq(PLD_IRQ_CFIREQ);
 
 #if defined(CONFIG_M32R_CFC)
 	/* ICUCR41: CFC Insert */
-	irq = PLD_IRQ_CFC_INSERT;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_CFC_INSERT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_INSERT].chip = &mappi2_irq_type;
+	irq_desc[PLD_IRQ_CFC_INSERT].action = 0;
+	irq_desc[PLD_IRQ_CFC_INSERT].depth = 1;	/* disable nested irq */
 	icu_data[PLD_IRQ_CFC_INSERT].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD00;
 	disable_mappi2_irq(PLD_IRQ_CFC_INSERT);
 
 	/* ICUCR42: CFC Eject */
-	irq = PLD_IRQ_CFC_EJECT;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi2_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_CFC_EJECT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_EJECT].chip = &mappi2_irq_type;
+	irq_desc[PLD_IRQ_CFC_EJECT].action = 0;
+	irq_desc[PLD_IRQ_CFC_EJECT].depth = 1;	/* disable nested irq */
 	icu_data[PLD_IRQ_CFC_EJECT].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_mappi2_irq(PLD_IRQ_CFC_EJECT);
 #endif /* CONFIG_MAPPI2_CFC */
Index: linux-2.6/arch/m32r/platforms/mappi3/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/mappi3/setup.c
+++ linux-2.6/arch/m32r/platforms/mappi3/setup.c
@@ -83,114 +83,91 @@ static struct hw_interrupt_type mappi3_i
 
 void __init init_IRQ(void)
 {
-	struct irq_desc *desc;
-	unsigned int irq;
-
 #if defined(CONFIG_SMC91X)
 	/* INT0 : LAN controller (SMC91111) */
-	irq = M32R_IRQ_INT0;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT0].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT0].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_INT0].action = 0;
+	irq_desc[M32R_IRQ_INT0].depth = 1;
 	icu_data[M32R_IRQ_INT0].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_mappi3_irq(M32R_IRQ_INT0);
 #endif  /* CONFIG_SMC91X */
 
 	/* MFT2 : system timer */
-	irq = M32R_IRQ_MFT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_mappi3_irq(M32R_IRQ_MFT2);
 
 #ifdef CONFIG_SERIAL_M32R_SIO
 	/* SIO0_R : uart receive data */
-	irq = M32R_IRQ_SIO0_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_mappi3_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0_S : uart send data */
-	irq = M32R_IRQ_SIO0_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_mappi3_irq(M32R_IRQ_SIO0_S);
 	/* SIO1_R : uart receive data */
-	irq = M32R_IRQ_SIO1_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = 0;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_mappi3_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1_S : uart send data */
-	irq = M32R_IRQ_SIO1_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = 0;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_mappi3_irq(M32R_IRQ_SIO1_S);
 #endif  /* CONFIG_M32R_USE_DBG_CONSOLE */
 
 #if defined(CONFIG_USB)
 	/* INT1 : USB Host controller interrupt */
-	irq = M32R_IRQ_INT1;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT1].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT1].chip = &mappi3_irq_type;
+	irq_desc[M32R_IRQ_INT1].action = 0;
+	irq_desc[M32R_IRQ_INT1].depth = 1;
 	icu_data[M32R_IRQ_INT1].icucr = M32R_ICUCR_ISMOD01;
 	disable_mappi3_irq(M32R_IRQ_INT1);
 #endif /* CONFIG_USB */
 
 	/* CFC IREQ */
-	irq = PLD_IRQ_CFIREQ;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_CFIREQ].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFIREQ].chip = &mappi3_irq_type;
+	irq_desc[PLD_IRQ_CFIREQ].action = 0;
+	irq_desc[PLD_IRQ_CFIREQ].depth = 1;	/* disable nested irq */
 	icu_data[PLD_IRQ_CFIREQ].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD01;
 	disable_mappi3_irq(PLD_IRQ_CFIREQ);
 
 #if defined(CONFIG_M32R_CFC)
 	/* ICUCR41: CFC Insert & eject */
-	irq = PLD_IRQ_CFC_INSERT;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_CFC_INSERT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_INSERT].chip = &mappi3_irq_type;
+	irq_desc[PLD_IRQ_CFC_INSERT].action = 0;
+	irq_desc[PLD_IRQ_CFC_INSERT].depth = 1;	/* disable nested irq */
 	icu_data[PLD_IRQ_CFC_INSERT].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD00;
 	disable_mappi3_irq(PLD_IRQ_CFC_INSERT);
 
 #endif /* CONFIG_M32R_CFC */
 
 	/* IDE IREQ */
-	irq = PLD_IRQ_IDEIREQ;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi3_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_IDEIREQ].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_IDEIREQ].chip = &mappi3_irq_type;
+	irq_desc[PLD_IRQ_IDEIREQ].action = 0;
+	irq_desc[PLD_IRQ_IDEIREQ].depth = 1;	/* disable nested irq */
 	icu_data[PLD_IRQ_IDEIREQ].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_mappi3_irq(PLD_IRQ_IDEIREQ);
 
Index: linux-2.6/arch/m32r/platforms/oaks32r/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/oaks32r/setup.c
+++ linux-2.6/arch/m32r/platforms/oaks32r/setup.c
@@ -75,8 +75,6 @@ static struct hw_interrupt_type oaks32r_
 void __init init_IRQ(void)
 {
 	static int once = 0;
-	struct irq_desc *desc;
-	unsigned int irq;
 
 	if (once)
 		return;
@@ -85,64 +83,52 @@ void __init init_IRQ(void)
 
 #ifdef CONFIG_NE2000
 	/* INT3 : LAN controller (RTL8019AS) */
-	irq = M32R_IRQ_INT3;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &oaks32r_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_INT3].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT3].chip = &oaks32r_irq_type;
+	irq_desc[M32R_IRQ_INT3].action = 0;
+	irq_desc[M32R_IRQ_INT3].depth = 1;
 	icu_data[M32R_IRQ_INT3].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_oaks32r_irq(M32R_IRQ_INT3);
 #endif /* CONFIG_M32R_NE2000 */
 
 	/* MFT2 : system timer */
-	irq = M32R_IRQ_MFT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &oaks32r_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &oaks32r_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_oaks32r_irq(M32R_IRQ_MFT2);
 
 #ifdef CONFIG_SERIAL_M32R_SIO
 	/* SIO0_R : uart receive data */
-	irq = M32R_IRQ_SIO0_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &oaks32r_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &oaks32r_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_oaks32r_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0_S : uart send data */
-	irq = M32R_IRQ_SIO0_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &oaks32r_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &oaks32r_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_oaks32r_irq(M32R_IRQ_SIO0_S);
 
 	/* SIO1_R : uart receive data */
-	irq = M32R_IRQ_SIO1_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &oaks32r_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &oaks32r_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = 0;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_oaks32r_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1_S : uart send data */
-	irq = M32R_IRQ_SIO1_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &oaks32r_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &oaks32r_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = 0;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_oaks32r_irq(M32R_IRQ_SIO1_S);
 #endif /* CONFIG_SERIAL_M32R_SIO */
Index: linux-2.6/arch/m32r/platforms/opsput/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/opsput/setup.c
+++ linux-2.6/arch/m32r/platforms/opsput/setup.c
@@ -298,130 +298,103 @@ static struct hw_interrupt_type opsput_l
 
 void __init init_IRQ(void)
 {
-	struct irq_desc *desc;
-	unsigned int irq;
-
 #if defined(CONFIG_SMC91X)
 	/* INT#0: LAN controller on OPSPUT-LAN (SMC91C111)*/
-	irq = OPSPUT_LAN_IRQ_LAN;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[OPSPUT_LAN_IRQ_LAN].status = IRQ_DISABLED;
+	irq_desc[OPSPUT_LAN_IRQ_LAN].chip = &opsput_lanpld_irq_type;
+	irq_desc[OPSPUT_LAN_IRQ_LAN].action = 0;
+	irq_desc[OPSPUT_LAN_IRQ_LAN].depth = 1;	/* disable nested irq */
 	lanpld_icu_data[irq2lanpldirq(OPSPUT_LAN_IRQ_LAN)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD02;	/* "H" edge sense */
 	disable_opsput_lanpld_irq(OPSPUT_LAN_IRQ_LAN);
 #endif  /* CONFIG_SMC91X */
 
 	/* MFT2 : system timer */
-	irq = M32R_IRQ_MFT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_opsput_irq(M32R_IRQ_MFT2);
 
 	/* SIO0 : receive */
-	irq = M32R_IRQ_SIO0_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_opsput_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0 : send */
-	irq = M32R_IRQ_SIO0_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_opsput_irq(M32R_IRQ_SIO0_S);
 
 	/* SIO1 : receive */
-	irq = M32R_IRQ_SIO1_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = 0;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_opsput_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1 : send */
-	irq = M32R_IRQ_SIO1_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = 0;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_opsput_irq(M32R_IRQ_SIO1_S);
 
 	/* DMA1 : */
-	irq = M32R_IRQ_DMA1;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_DMA1].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_DMA1].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_DMA1].action = 0;
+	irq_desc[M32R_IRQ_DMA1].depth = 1;
 	icu_data[M32R_IRQ_DMA1].icucr = 0;
 	disable_opsput_irq(M32R_IRQ_DMA1);
 
 #ifdef CONFIG_SERIAL_M32R_PLDSIO
 	/* INT#1: SIO0 Receive on PLD */
-	irq = PLD_IRQ_SIO0_RCV;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_SIO0_RCV].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_SIO0_RCV].chip = &opsput_pld_irq_type;
+	irq_desc[PLD_IRQ_SIO0_RCV].action = 0;
+	irq_desc[PLD_IRQ_SIO0_RCV].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_SIO0_RCV)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD03;
 	disable_opsput_pld_irq(PLD_IRQ_SIO0_RCV);
 
 	/* INT#1: SIO0 Send on PLD */
-	irq = PLD_IRQ_SIO0_SND;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_SIO0_SND].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_SIO0_SND].chip = &opsput_pld_irq_type;
+	irq_desc[PLD_IRQ_SIO0_SND].action = 0;
+	irq_desc[PLD_IRQ_SIO0_SND].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_SIO0_SND)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD03;
 	disable_opsput_pld_irq(PLD_IRQ_SIO0_SND);
 #endif  /* CONFIG_SERIAL_M32R_PLDSIO */
 
 	/* INT#1: CFC IREQ on PLD */
-	irq = PLD_IRQ_CFIREQ;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_CFIREQ].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFIREQ].chip = &opsput_pld_irq_type;
+	irq_desc[PLD_IRQ_CFIREQ].action = 0;
+	irq_desc[PLD_IRQ_CFIREQ].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_CFIREQ)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD01;	/* 'L' level sense */
 	disable_opsput_pld_irq(PLD_IRQ_CFIREQ);
 
 	/* INT#1: CFC Insert on PLD */
-	irq = PLD_IRQ_CFC_INSERT;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_CFC_INSERT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_INSERT].chip = &opsput_pld_irq_type;
+	irq_desc[PLD_IRQ_CFC_INSERT].action = 0;
+	irq_desc[PLD_IRQ_CFC_INSERT].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_CFC_INSERT)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD00;	/* 'L' edge sense */
 	disable_opsput_pld_irq(PLD_IRQ_CFC_INSERT);
 
 	/* INT#1: CFC Eject on PLD */
-	irq = PLD_IRQ_CFC_EJECT;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[PLD_IRQ_CFC_EJECT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_CFC_EJECT].chip = &opsput_pld_irq_type;
+	irq_desc[PLD_IRQ_CFC_EJECT].action = 0;
+	irq_desc[PLD_IRQ_CFC_EJECT].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_CFC_EJECT)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD02;	/* 'H' edge sense */
 	disable_opsput_pld_irq(PLD_IRQ_CFC_EJECT);
 
@@ -436,20 +409,18 @@ void __init init_IRQ(void)
 	 * INT1# is used for UART, MMC, CF Controller in FPGA.
 	 * We enable it here.
 	 */
-	icr_data[M32R_IRQ_INT1].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD11;
+	icu_data[M32R_IRQ_INT1].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD11;
 	enable_opsput_irq(M32R_IRQ_INT1);
 
 #if defined(CONFIG_USB)
 	outw(USBCR_OTGS, USBCR); 	/* USBCR: non-OTG */
 
-	irq = OPSPUT_LCD_IRQ_USB_INT1;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
-	lcdpld_icu_data[irq2lcdpldirq(OPSPUT_LCD_IRQ_USB_INT1)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD01;	/* "L" level sense */
-	disable_opsput_lcdpld_irq(OPSPUT_LCD_IRQ_USB_INT1);
+    irq_desc[OPSPUT_LCD_IRQ_USB_INT1].status = IRQ_DISABLED;
+    irq_desc[OPSPUT_LCD_IRQ_USB_INT1].chip = &opsput_lcdpld_irq_type;
+    irq_desc[OPSPUT_LCD_IRQ_USB_INT1].action = 0;
+    irq_desc[OPSPUT_LCD_IRQ_USB_INT1].depth = 1;
+    lcdpld_icu_data[irq2lcdpldirq(OPSPUT_LCD_IRQ_USB_INT1)].icucr = PLD_ICUCR_IEN|PLD_ICUCR_ISMOD01;	/* "L" level sense */
+    disable_opsput_lcdpld_irq(OPSPUT_LCD_IRQ_USB_INT1);
 #endif
 	/*
 	 * INT2# is used for BAT, USB, AUDIO
@@ -462,12 +433,10 @@ void __init init_IRQ(void)
 	/*
 	 * INT3# is used for AR
 	 */
-	irq = M32R_IRQ_INT3;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &opsput_lanpld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;	/* disable nested irq */
+	irq_desc[M32R_IRQ_INT3].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_INT3].chip = &opsput_irq_type;
+	irq_desc[M32R_IRQ_INT3].action = 0;
+	irq_desc[M32R_IRQ_INT3].depth = 1;
 	icu_data[M32R_IRQ_INT3].icucr = M32R_ICUCR_IEN|M32R_ICUCR_ISMOD10;
 	disable_opsput_irq(M32R_IRQ_INT3);
 #endif /* CONFIG_VIDEO_M32R_AR */
Index: linux-2.6/arch/m32r/platforms/usrv/setup.c
===================================================================
--- linux-2.6.orig/arch/m32r/platforms/usrv/setup.c
+++ linux-2.6/arch/m32r/platforms/usrv/setup.c
@@ -149,8 +149,6 @@ void __init init_IRQ(void)
 {
 	static int once = 0;
 	int i;
-	struct irq_desc *desc;
-	unsigned int irq;
 
 	if (once)
 		return;
@@ -158,65 +156,53 @@ void __init init_IRQ(void)
 		once++;
 
 	/* MFT2 : system timer */
-	irq = M32R_IRQ_MFT2;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_MFT2].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_MFT2].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_MFT2].action = 0;
+	irq_desc[M32R_IRQ_MFT2].depth = 1;
 	icu_data[M32R_IRQ_MFT2].icucr = M32R_ICUCR_IEN;
 	disable_mappi_irq(M32R_IRQ_MFT2);
 
 #if defined(CONFIG_SERIAL_M32R_SIO)
 	/* SIO0_R : uart receive data */
-	irq = M32R_IRQ_SIO0_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_R].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO0_R].action = 0;
+	irq_desc[M32R_IRQ_SIO0_R].depth = 1;
 	icu_data[M32R_IRQ_SIO0_R].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO0_R);
 
 	/* SIO0_S : uart send data */
-	irq = M32R_IRQ_SIO0_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO0_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO0_S].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO0_S].action = 0;
+	irq_desc[M32R_IRQ_SIO0_S].depth = 1;
 	icu_data[M32R_IRQ_SIO0_S].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO0_S);
 
 	/* SIO1_R : uart receive data */
-	irq = M32R_IRQ_SIO1_R;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_R].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_R].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO1_R].action = 0;
+	irq_desc[M32R_IRQ_SIO1_R].depth = 1;
 	icu_data[M32R_IRQ_SIO1_R].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO1_R);
 
 	/* SIO1_S : uart send data */
-	irq = M32R_IRQ_SIO1_S;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &mappi_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[M32R_IRQ_SIO1_S].status = IRQ_DISABLED;
+	irq_desc[M32R_IRQ_SIO1_S].chip = &mappi_irq_type;
+	irq_desc[M32R_IRQ_SIO1_S].action = 0;
+	irq_desc[M32R_IRQ_SIO1_S].depth = 1;
 	icu_data[M32R_IRQ_SIO1_S].icucr = 0;
 	disable_mappi_irq(M32R_IRQ_SIO1_S);
 #endif  /* CONFIG_SERIAL_M32R_SIO */
 
 	/* INT#67-#71: CFC#0 IREQ on PLD */
 	for (i = 0 ; i < CONFIG_M32R_CFC_NUM ; i++ ) {
-		irq = PLD_IRQ_CF0 + i;
-		desc = irq_desc(irq);
-		desc->status = IRQ_DISABLED;
-		desc->chip = &m32700ut_pld_irq_type;
-		desc->action = 0;
-		desc->depth = 1;
+		irq_desc[PLD_IRQ_CF0 + i].status = IRQ_DISABLED;
+		irq_desc[PLD_IRQ_CF0 + i].chip = &m32700ut_pld_irq_type;
+		irq_desc[PLD_IRQ_CF0 + i].action = 0;
+		irq_desc[PLD_IRQ_CF0 + i].depth = 1;	/* disable nested irq */
 		pld_icu_data[irq2pldirq(PLD_IRQ_CF0 + i)].icucr
 			= PLD_ICUCR_ISMOD01;	/* 'L' level sense */
 		disable_m32700ut_pld_irq(PLD_IRQ_CF0 + i);
@@ -224,23 +210,19 @@ void __init init_IRQ(void)
 
 #if defined(CONFIG_SERIAL_8250) || defined(CONFIG_SERIAL_8250_MODULE)
 	/* INT#76: 16552D#0 IREQ on PLD */
-	irq = PLD_IRQ_UART0;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_UART0].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_UART0].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_UART0].action = 0;
+	irq_desc[PLD_IRQ_UART0].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_UART0)].icucr
 		= PLD_ICUCR_ISMOD03;	/* 'H' level sense */
 	disable_m32700ut_pld_irq(PLD_IRQ_UART0);
 
 	/* INT#77: 16552D#1 IREQ on PLD */
-	irq = PLD_IRQ_UART1;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_UART1].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_UART1].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_UART1].action = 0;
+	irq_desc[PLD_IRQ_UART1].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_UART1)].icucr
 		= PLD_ICUCR_ISMOD03;	/* 'H' level sense */
 	disable_m32700ut_pld_irq(PLD_IRQ_UART1);
@@ -248,12 +230,10 @@ void __init init_IRQ(void)
 
 #if defined(CONFIG_IDC_AK4524) || defined(CONFIG_IDC_AK4524_MODULE)
 	/* INT#80: AK4524 IREQ on PLD */
-	irq = PLD_IRQ_SNDINT;
-	desc = irq_desc(irq);
-	desc->status = IRQ_DISABLED;
-	desc->chip = &m32700ut_pld_irq_type;
-	desc->action = 0;
-	desc->depth = 1;
+	irq_desc[PLD_IRQ_SNDINT].status = IRQ_DISABLED;
+	irq_desc[PLD_IRQ_SNDINT].chip = &m32700ut_pld_irq_type;
+	irq_desc[PLD_IRQ_SNDINT].action = 0;
+	irq_desc[PLD_IRQ_SNDINT].depth = 1;	/* disable nested irq */
 	pld_icu_data[irq2pldirq(PLD_IRQ_SNDINT)].icucr
 		= PLD_ICUCR_ISMOD01;	/* 'L' level sense */
 	disable_m32700ut_pld_irq(PLD_IRQ_SNDINT);
Index: linux-2.6/arch/m68knommu/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m68knommu/kernel/irq.c
+++ linux-2.6/arch/m68knommu/kernel/irq.c
@@ -23,7 +23,7 @@ asmlinkage void do_IRQ(int irq, struct p
 	struct pt_regs *oldregs = set_irq_regs(regs);
 
 	irq_enter();
-	__do_IRQ(irq, irq_desc(irq));
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 
 	set_irq_regs(oldregs);
@@ -48,12 +48,10 @@ void __init init_IRQ(void)
 	init_vectors();
 
 	for (irq = 0; (irq < NR_IRQS); irq++) {
-		struct irq_desc *desc = irq_desc(irq);
-
-		desc->status = IRQ_DISABLED;
-		desc->action = NULL;
-		desc->depth = 1;
-		desc->chip = &m_irq_chip;
+		irq_desc[irq].status = IRQ_DISABLED;
+		irq_desc[irq].action = NULL;
+		irq_desc[irq].depth = 1;
+		irq_desc[irq].chip = &m_irq_chip;
 	}
 }
 
@@ -66,13 +64,11 @@ int show_interrupts(struct seq_file *p,
 		seq_puts(p, "           CPU0\n");
 
 	if (irq < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(irq);
-
-		ap = desc->action;
+		ap = irq_desc[irq].action;
 		if (ap) {
 			seq_printf(p, "%3d: ", irq);
 			seq_printf(p, "%10u ", kstat_irqs(irq));
-			seq_printf(p, "%14s  ", desc->chip->name);
+			seq_printf(p, "%14s  ", irq_desc[irq].chip->name);
 
 			seq_printf(p, "%s", ap->name);
 			for (ap = ap->next; ap; ap = ap->next)
Index: linux-2.6/arch/mips/au1000/common/irq.c
===================================================================
--- linux-2.6.orig/arch/mips/au1000/common/irq.c
+++ linux-2.6/arch/mips/au1000/common/irq.c
@@ -224,7 +224,7 @@ static inline void mask_and_ack_level_ir
 
 static void end_irq(unsigned int irq_nr)
 {
-	if (!(irq_desc(irq_nr)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq_nr].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		local_enable_irq(irq_nr);
 
 #if defined(CONFIG_MIPS_PB1000)
Index: linux-2.6/arch/mips/dec/ioasic-irq.c
===================================================================
--- linux-2.6.orig/arch/mips/dec/ioasic-irq.c
+++ linux-2.6/arch/mips/dec/ioasic-irq.c
@@ -57,7 +57,7 @@ static inline void ack_ioasic_irq(unsign
 
 static inline void end_ioasic_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		unmask_ioasic_irq(irq);
 }
 
Index: linux-2.6/arch/mips/emma2rh/markeins/irq_markeins.c
===================================================================
--- linux-2.6.orig/arch/mips/emma2rh/markeins/irq_markeins.c
+++ linux-2.6/arch/mips/emma2rh/markeins/irq_markeins.c
@@ -110,7 +110,7 @@ static void emma2rh_gpio_irq_ack(unsigne
 
 static void emma2rh_gpio_irq_end(unsigned int irq)
 {
-	if (!(irq_desc(irq)status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		ll_emma2rh_gpio_irq_enable(irq - emma2rh_gpio_irq_base);
 }
 
Index: linux-2.6/arch/mips/kernel/irq-gic.c
===================================================================
--- linux-2.6.orig/arch/mips/kernel/irq-gic.c
+++ linux-2.6/arch/mips/kernel/irq-gic.c
@@ -187,7 +187,7 @@ static void gic_set_affinity(unsigned in
 		set_bit(irq, pcpu_masks[first_cpu(tmp)].pcpu_mask);
 
 	}
-	irq_desc(irq)->affinity = cpumask;
+	irq_desc[irq].affinity = cpumask;
 	spin_unlock_irqrestore(&gic_lock, flags);
 
 }
Index: linux-2.6/arch/mips/kernel/irq-msc01.c
===================================================================
--- linux-2.6.orig/arch/mips/kernel/irq-msc01.c
+++ linux-2.6/arch/mips/kernel/irq-msc01.c
@@ -79,7 +79,7 @@ static void edge_mask_and_ack_msc_irq(un
  */
 static void end_msc_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		unmask_msc_irq(irq);
 }
 
Index: linux-2.6/arch/mips/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/mips/kernel/irq.c
+++ linux-2.6/arch/mips/kernel/irq.c
@@ -99,10 +99,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ", i);
@@ -112,7 +110,7 @@ int show_interrupts(struct seq_file *p,
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
-		seq_printf(p, " %14s", desc->chip->name);
+		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -120,7 +118,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 		seq_putc(p, '\n');
 		seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
Index: linux-2.6/arch/mips/kernel/smtc.c
===================================================================
--- linux-2.6.orig/arch/mips/kernel/smtc.c
+++ linux-2.6/arch/mips/kernel/smtc.c
@@ -663,7 +663,7 @@ void smtc_forward_irq(unsigned int irq)
 	 * and efficiency, we just pick the easiest one to find.
 	 */
 
-	target = first_cpu(irq_desc(irq)->affinity);
+	target = first_cpu(irq_desc[irq].affinity);
 
 	/*
 	 * We depend on the platform code to have correctly processed
Index: linux-2.6/arch/mips/mti-malta/malta-smtc.c
===================================================================
--- linux-2.6.orig/arch/mips/mti-malta/malta-smtc.c
+++ linux-2.6/arch/mips/mti-malta/malta-smtc.c
@@ -138,7 +138,7 @@ void plat_set_irq_affinity(unsigned int
 		if ((cpu_data[cpu].vpe_id != 0) || !cpu_online(cpu))
 			cpu_clear(cpu, tmask);
 	}
-	irq_desc(irq)->affinity = tmask;
+	irq_desc[irq].affinity = tmask;
 
 	if (cpus_empty(tmask))
 		/*
Index: linux-2.6/arch/mips/sgi-ip32/ip32-irq.c
===================================================================
--- linux-2.6.orig/arch/mips/sgi-ip32/ip32-irq.c
+++ linux-2.6/arch/mips/sgi-ip32/ip32-irq.c
@@ -157,7 +157,7 @@ static void crime_level_mask_and_ack_irq
 
 static void crime_level_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq).status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		crime_enable_irq(irq);
 }
 
@@ -186,7 +186,7 @@ static void crime_edge_mask_and_ack_irq(
 
 static void crime_edge_end_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		crime_enable_irq(irq);
 }
 
@@ -227,7 +227,7 @@ static void disable_macepci_irq(unsigned
 
 static void end_macepci_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_macepci_irq(irq);
 }
 
@@ -340,7 +340,7 @@ static void mask_and_ack_maceisa_irq(uns
 
 static void end_maceisa_irq(unsigned irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS)))
 		enable_maceisa_irq(irq);
 }
 
@@ -376,7 +376,7 @@ static void disable_mace_irq(unsigned in
 
 static void end_mace_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_mace_irq(irq);
 }
 
Index: linux-2.6/arch/mips/sibyte/bcm1480/irq.c
===================================================================
--- linux-2.6.orig/arch/mips/sibyte/bcm1480/irq.c
+++ linux-2.6/arch/mips/sibyte/bcm1480/irq.c
@@ -113,7 +113,7 @@ static void bcm1480_set_affinity(unsigne
 {
 	int i = 0, old_cpu, cpu, int_on, k;
 	u64 cur_ints;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	unsigned long flags;
 	unsigned int irq_dirty;
 
@@ -225,8 +225,9 @@ static void ack_bcm1480_irq(unsigned int
 
 static void end_bcm1480_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS))) {
 		bcm1480_unmask_irq(bcm1480_irq_owner[irq], irq);
+	}
 }
 
 
Index: linux-2.6/arch/mips/sibyte/sb1250/irq.c
===================================================================
--- linux-2.6.orig/arch/mips/sibyte/sb1250/irq.c
+++ linux-2.6/arch/mips/sibyte/sb1250/irq.c
@@ -107,7 +107,7 @@ static void sb1250_set_affinity(unsigned
 {
 	int i = 0, old_cpu, cpu, int_on;
 	u64 cur_ints;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	unsigned long flags;
 
 	i = first_cpu(mask);
@@ -209,8 +209,9 @@ static void ack_sb1250_irq(unsigned int
 
 static void end_sb1250_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS))) {
 		sb1250_unmask_irq(sb1250_irq_owner[irq], irq);
+	}
 }
 
 
Index: linux-2.6/arch/mips/sni/a20r.c
===================================================================
--- linux-2.6.orig/arch/mips/sni/a20r.c
+++ linux-2.6/arch/mips/sni/a20r.c
@@ -181,7 +181,7 @@ static inline void mask_a20r_irq(unsigne
 
 static void end_a20r_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED | IRQ_INPROGRESS))) {
+	if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS))) {
 		a20r_ack_hwint();
 		unmask_a20r_irq(irq);
 	}
Index: linux-2.6/arch/mips/sni/pcimt.c
===================================================================
--- linux-2.6.orig/arch/mips/sni/pcimt.c
+++ linux-2.6/arch/mips/sni/pcimt.c
@@ -209,7 +209,7 @@ void disable_pcimt_irq(unsigned int irq)
 
 static void end_pcimt_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_pcimt_irq(irq);
 }
 
Index: linux-2.6/arch/mips/sni/pcit.c
===================================================================
--- linux-2.6.orig/arch/mips/sni/pcit.c
+++ linux-2.6/arch/mips/sni/pcit.c
@@ -171,7 +171,7 @@ void disable_pcit_irq(unsigned int irq)
 
 void end_pcit_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_pcit_irq(irq);
 }
 
Index: linux-2.6/arch/mips/sni/rm200.c
===================================================================
--- linux-2.6.orig/arch/mips/sni/rm200.c
+++ linux-2.6/arch/mips/sni/rm200.c
@@ -443,7 +443,7 @@ void disable_rm200_irq(unsigned int irq)
 
 void end_rm200_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_rm200_irq(irq);
 }
 
Index: linux-2.6/arch/mips/vr41xx/common/icu.c
===================================================================
--- linux-2.6.orig/arch/mips/vr41xx/common/icu.c
+++ linux-2.6/arch/mips/vr41xx/common/icu.c
@@ -154,7 +154,7 @@ static inline uint16_t icu2_clear(uint8_
 
 void vr41xx_enable_piuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(PIU_IRQ);
+	struct irq_desc *desc = irq_desc + PIU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4111 ||
@@ -169,7 +169,7 @@ EXPORT_SYMBOL(vr41xx_enable_piuint);
 
 void vr41xx_disable_piuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(PIU_IRQ);
+	struct irq_desc *desc = irq_desc + PIU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4111 ||
@@ -184,7 +184,7 @@ EXPORT_SYMBOL(vr41xx_disable_piuint);
 
 void vr41xx_enable_aiuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(AIU_IRQ);
+	struct irq_desc *desc = irq_desc + AIU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4111 ||
@@ -199,7 +199,7 @@ EXPORT_SYMBOL(vr41xx_enable_aiuint);
 
 void vr41xx_disable_aiuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(AIU_IRQ);
+	struct irq_desc *desc = irq_desc + AIU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4111 ||
@@ -214,7 +214,7 @@ EXPORT_SYMBOL(vr41xx_disable_aiuint);
 
 void vr41xx_enable_kiuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(KIU_IRQ);
+	struct irq_desc *desc = irq_desc + KIU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4111 ||
@@ -229,7 +229,7 @@ EXPORT_SYMBOL(vr41xx_enable_kiuint);
 
 void vr41xx_disable_kiuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(KIU_IRQ);
+	struct irq_desc *desc = irq_desc + KIU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4111 ||
@@ -244,7 +244,7 @@ EXPORT_SYMBOL(vr41xx_disable_kiuint);
 
 void vr41xx_enable_macint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(ETHERNET_IRQ);
+	struct irq_desc *desc = irq_desc + ETHERNET_IRQ;
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
@@ -256,7 +256,7 @@ EXPORT_SYMBOL(vr41xx_enable_macint);
 
 void vr41xx_disable_macint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(ETHERNET_IRQ);
+	struct irq_desc *desc = irq_desc + ETHERNET_IRQ;
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
@@ -268,7 +268,7 @@ EXPORT_SYMBOL(vr41xx_disable_macint);
 
 void vr41xx_enable_dsiuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(DSIU_IRQ);
+	struct irq_desc *desc = irq_desc + DSIU_IRQ;
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
@@ -280,7 +280,7 @@ EXPORT_SYMBOL(vr41xx_enable_dsiuint);
 
 void vr41xx_disable_dsiuint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(DSIU_IRQ);
+	struct irq_desc *desc = irq_desc + DSIU_IRQ;
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
@@ -292,7 +292,7 @@ EXPORT_SYMBOL(vr41xx_disable_dsiuint);
 
 void vr41xx_enable_firint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(FIR_IRQ);
+	struct irq_desc *desc = irq_desc + FIR_IRQ;
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
@@ -304,7 +304,7 @@ EXPORT_SYMBOL(vr41xx_enable_firint);
 
 void vr41xx_disable_firint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(FIR_IRQ);
+	struct irq_desc *desc = irq_desc + FIR_IRQ;
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
@@ -316,7 +316,7 @@ EXPORT_SYMBOL(vr41xx_disable_firint);
 
 void vr41xx_enable_pciint(void)
 {
-	struct irq_desc *desc = irq_desc(PCI_IRQ);
+	struct irq_desc *desc = irq_desc + PCI_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -332,7 +332,7 @@ EXPORT_SYMBOL(vr41xx_enable_pciint);
 
 void vr41xx_disable_pciint(void)
 {
-	struct irq_desc *desc = irq_desc(PCI_IRQ);
+	struct irq_desc *desc = irq_desc + PCI_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -348,7 +348,7 @@ EXPORT_SYMBOL(vr41xx_disable_pciint);
 
 void vr41xx_enable_scuint(void)
 {
-	struct irq_desc *desc = irq_desc(SCU_IRQ);
+	struct irq_desc *desc = irq_desc + SCU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -364,7 +364,7 @@ EXPORT_SYMBOL(vr41xx_enable_scuint);
 
 void vr41xx_disable_scuint(void)
 {
-	struct irq_desc *desc = irq_desc(SCU_IRQ);
+	struct irq_desc *desc = irq_desc + SCU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -380,7 +380,7 @@ EXPORT_SYMBOL(vr41xx_disable_scuint);
 
 void vr41xx_enable_csiint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(CSI_IRQ);
+	struct irq_desc *desc = irq_desc + CSI_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -396,7 +396,7 @@ EXPORT_SYMBOL(vr41xx_enable_csiint);
 
 void vr41xx_disable_csiint(uint16_t mask)
 {
-	struct irq_desc *desc = irq_desc(CSI_IRQ);
+	struct irq_desc *desc = irq_desc + CSI_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -412,7 +412,7 @@ EXPORT_SYMBOL(vr41xx_disable_csiint);
 
 void vr41xx_enable_bcuint(void)
 {
-	struct irq_desc *desc = irq_desc(BCU_IRQ);
+	struct irq_desc *desc = irq_desc + BCU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -428,7 +428,7 @@ EXPORT_SYMBOL(vr41xx_enable_bcuint);
 
 void vr41xx_disable_bcuint(void)
 {
-	struct irq_desc *desc = irq_desc(BCU_IRQ);
+	struct irq_desc *desc = irq_desc + BCU_IRQ;
 	unsigned long flags;
 
 	if (current_cpu_type() == CPU_VR4122 ||
@@ -480,7 +480,7 @@ static struct irq_chip sysint2_irq_type
 
 static inline int set_sysint1_assign(unsigned int irq, unsigned char assign)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	uint16_t intassign0, intassign1;
 	unsigned int pin;
 
@@ -540,7 +540,7 @@ static inline int set_sysint1_assign(uns
 
 static inline int set_sysint2_assign(unsigned int irq, unsigned char assign)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 	uint16_t intassign2, intassign3;
 	unsigned int pin;
 
Index: linux-2.6/arch/mips/vr41xx/common/irq.c
===================================================================
--- linux-2.6.orig/arch/mips/vr41xx/common/irq.c
+++ linux-2.6/arch/mips/vr41xx/common/irq.c
@@ -72,7 +72,7 @@ static void irq_dispatch(unsigned int ir
 	cascade = irq_cascade + irq;
 	if (cascade->get_irq != NULL) {
 		unsigned int source_irq = irq;
-		desc = irq_desc(source_irq);
+		desc = irq_desc + source_irq;
 		if (desc->chip->mask_ack)
 			desc->chip->mask_ack(source_irq);
 		else {
Index: linux-2.6/arch/mn10300/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/mn10300/kernel/irq.c
+++ linux-2.6/arch/mn10300/kernel/irq.c
@@ -125,7 +125,7 @@ void __init init_IRQ(void)
 	int irq;
 
 	for (irq = 0; irq < NR_IRQS; irq++)
-		if (irq_desc(irq)->chip == &no_irq_type)
+		if (irq_desc[irq].chip == &no_irq_type)
 			set_irq_chip_and_handler(irq, &mn10300_cpu_pic,
 						 handle_edge_irq);
 	unit_init_IRQ();
@@ -183,7 +183,6 @@ int show_interrupts(struct seq_file *p,
 	int i = *(loff_t *) v, j, cpu;
 	struct irqaction *action;
 	unsigned long flags;
-	struct irq_desc *desc;
 
 	switch (i) {
 		/* display column title bar naming CPUs */
@@ -197,15 +196,13 @@ int show_interrupts(struct seq_file *p,
 
 		/* display information rows, one per active CPU */
 	case 1 ... NR_IRQS - 1:
-		desc = irq_desc(i);
-		spin_lock_irqsave(&desc->lock, flags);
-
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (action) {
 			seq_printf(p, "%3d: ", i);
 			for_each_present_cpu(cpu)
 				seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
-			seq_printf(p, " %14s.%u", desc->chip->name,
+			seq_printf(p, " %14s.%u", irq_desc[i].chip->name,
 				   (GxICR(i) & GxICR_LEVEL) >>
 				   GxICR_LEVEL_SHIFT);
 			seq_printf(p, "  %s", action->name);
@@ -218,7 +215,7 @@ int show_interrupts(struct seq_file *p,
 			seq_putc(p, '\n');
 		}
 
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 		break;
 
 		/* polish off with NMI and error counters */
Index: linux-2.6/arch/parisc/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/parisc/kernel/irq.c
+++ linux-2.6/arch/parisc/kernel/irq.c
@@ -120,7 +120,7 @@ int cpu_check_affinity(unsigned int irq,
 	if (CHECK_IRQ_PER_CPU(irq)) {
 		/* Bad linux design decision.  The mask has already
 		 * been set; we must reset it */
-		irq_desc(irq)->affinity = CPU_MASK_ALL;
+		irq_desc[irq].affinity = CPU_MASK_ALL;
 		return -EINVAL;
 	}
 
@@ -136,7 +136,7 @@ static void cpu_set_affinity_irq(unsigne
 	if (cpu_check_affinity(irq, &dest))
 		return;
 
-	irq_desc(irq)->affinity = dest;
+	irq_desc[irq].affinity = dest;
 }
 #endif
 
@@ -175,10 +175,9 @@ int show_interrupts(struct seq_file *p,
 
 	if (i < NR_IRQS) {
 		struct irqaction *action;
-		struct irq_desc *desc = irq_desc(i);
 
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ", i);
@@ -189,7 +188,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #endif
 
-		seq_printf(p, " %14s", desc->chip->typename);
+		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 #ifndef PARISC_IRQ_CR16_COUNTS
 		seq_printf(p, "  %s", action->name);
 
@@ -221,7 +220,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
  skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	}
 
 	return 0;
@@ -239,16 +238,14 @@ int show_interrupts(struct seq_file *p,
 
 int cpu_claim_irq(unsigned int irq, struct irq_chip *type, void *data)
 {
-	struct irq_desc *desc = irq_desc(irq);
-
-	if (desc->action)
+	if (irq_desc[irq].action)
 		return -EBUSY;
-	if (desc->chip != &cpu_interrupt_type)
+	if (irq_desc[irq].chip != &cpu_interrupt_type)
 		return -EBUSY;
 
 	if (type) {
-		desc->chip = type;
-		desc->chip_data = data;
+		irq_desc[irq].chip = type;
+		irq_desc[irq].chip_data = data;
 		cpu_interrupt_type.enable(irq);
 	}
 	return 0;
@@ -298,7 +295,7 @@ int txn_alloc_irq(unsigned int bits_wide
 unsigned long txn_affinity_addr(unsigned int irq, int cpu)
 {
 #ifdef CONFIG_SMP
-	irq_desc(irq)->affinity = cpumask_of_cpu(cpu);
+	irq_desc[irq].affinity = cpumask_of_cpu(cpu);
 #endif
 
 	return cpu_data[cpu].txn_addr;
@@ -342,7 +339,6 @@ void do_cpu_irq_mask(struct pt_regs *reg
 	int irq, cpu = smp_processor_id();
 #ifdef CONFIG_SMP
 	cpumask_t dest;
-	struct irq_desc *desc;
 #endif
 
 	old_regs = set_irq_regs(regs);
@@ -355,9 +351,8 @@ void do_cpu_irq_mask(struct pt_regs *reg
 	irq = eirr_to_irq(eirr_val);
 
 #ifdef CONFIG_SMP
-	desc = irq_desc(irq);
-	dest = desc->affinity;
-	if (CHECK_IRQ_PER_CPU(desc->status) &&
+	dest = irq_desc[irq].affinity;
+	if (CHECK_IRQ_PER_CPU(irq_desc[irq].status) &&
 	    !cpu_isset(smp_processor_id(), dest)) {
 		int cpu = first_cpu(dest);
 
@@ -398,14 +393,14 @@ static void claim_cpu_irqs(void)
 {
 	int i;
 	for (i = CPU_IRQ_BASE; i <= CPU_IRQ_MAX; i++) {
-		irq_desc(i)->chip = &cpu_interrupt_type;
+		irq_desc[i].chip = &cpu_interrupt_type;
 	}
 
-	irq_desc(TIMER_IRQ)->action = &timer_action;
-	irq_desc(TIMER_IRQ)->status = IRQ_PER_CPU;
+	irq_desc[TIMER_IRQ].action = &timer_action;
+	irq_desc[TIMER_IRQ].status = IRQ_PER_CPU;
 #ifdef CONFIG_SMP
-	irq_desc(IPI_IRQ)->action = &ipi_action;
-	irq_desc(IPI_IRQ)->status = IRQ_PER_CPU;
+	irq_desc[IPI_IRQ].action = &ipi_action;
+	irq_desc[IPI_IRQ].status = IRQ_PER_CPU;
 #endif
 }
 
Index: linux-2.6/arch/powerpc/include/asm/irq.h
===================================================================
--- linux-2.6.orig/arch/powerpc/include/asm/irq.h
+++ linux-2.6/arch/powerpc/include/asm/irq.h
@@ -17,7 +17,7 @@
 #include <asm/atomic.h>
 
 
-#define get_irq_desc(irq) (irq_desc(irq))
+#define get_irq_desc(irq) (&irq_desc[(irq)])
 
 /* Define a way to iterate across irqs. */
 #define for_each_irq(i) \
Index: linux-2.6/arch/powerpc/kernel/crash.c
===================================================================
--- linux-2.6.orig/arch/powerpc/kernel/crash.c
+++ linux-2.6/arch/powerpc/kernel/crash.c
@@ -373,7 +373,7 @@ void default_machine_crash_shutdown(stru
 	hard_irq_disable();
 
 	for_each_irq(i) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = irq_desc + i;
 
 		if (desc->status & IRQ_INPROGRESS)
 			desc->chip->eoi(i);
Index: linux-2.6/arch/powerpc/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/powerpc/kernel/irq.c
+++ linux-2.6/arch/powerpc/kernel/irq.c
@@ -84,7 +84,7 @@ extern int tau_interrupts(int);
 #endif /* CONFIG_PPC32 */
 
 #ifdef CONFIG_PPC64
-EXPORT_SYMBOL(irq_descX);
+EXPORT_SYMBOL(irq_desc);
 
 int distribute_irqs = 1;
 
@@ -227,19 +227,18 @@ void fixup_irqs(cpumask_t map)
 
 	for_each_irq(irq) {
 		cpumask_t mask;
-		struct irq_desc *desc = irq_desc(irq);
 
-		if (desc->status & IRQ_PER_CPU)
+		if (irq_desc[irq].status & IRQ_PER_CPU)
 			continue;
 
-		cpus_and(mask, desc->affinity, map);
+		cpus_and(mask, irq_desc[irq].affinity, map);
 		if (any_online_cpu(mask) == NR_CPUS) {
 			printk("Breaking affinity for irq %i\n", irq);
 			mask = map;
 		}
-		if (desc->chip->set_affinity)
-			desc->chip->set_affinity(irq, mask);
-		else if (desc->action && !(warned++))
+		if (irq_desc[irq].chip->set_affinity)
+			irq_desc[irq].chip->set_affinity(irq, mask);
+		else if (irq_desc[irq].action && !(warned++))
 			printk("Cannot set affinity for irq %i\n", irq);
 	}
 
@@ -289,7 +288,7 @@ void do_IRQ(struct pt_regs *regs)
 		curtp = current_thread_info();
 		irqtp = hardirq_ctx[smp_processor_id()];
 		if (curtp != irqtp) {
-			struct irq_desc *desc = irq_desc(irq);
+			struct irq_desc *desc = irq_desc + irq;
 			void *handler = desc->handle_irq;
 			unsigned long saved_sp_limit = current->thread.ksp_limit;
 			if (handler == NULL)
Index: linux-2.6/arch/powerpc/platforms/8xx/m8xx_setup.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/8xx/m8xx_setup.c
+++ linux-2.6/arch/powerpc/platforms/8xx/m8xx_setup.c
@@ -223,7 +223,7 @@ static void cpm_cascade(unsigned int irq
 	int cascade_irq;
 
 	if ((cascade_irq = cpm_get_irq()) >= 0) {
-		struct irq_desc *cdesc = irq_desc(cascade_irq);
+		struct irq_desc *cdesc = irq_desc + cascade_irq;
 
 		generic_handle_irq(cascade_irq);
 		cdesc->chip->eoi(cascade_irq);
Index: linux-2.6/arch/powerpc/platforms/powermac/pic.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/powermac/pic.c
+++ linux-2.6/arch/powerpc/platforms/powermac/pic.c
@@ -153,7 +153,7 @@ static unsigned int pmac_startup_irq(uns
         int i = src >> 5;
 
   	spin_lock_irqsave(&pmac_pic_lock, flags);
-	if ((irq_desc(virq)->status & IRQ_LEVEL) == 0)
+	if ((irq_desc[virq].status & IRQ_LEVEL) == 0)
 		out_le32(&pmac_irq_hw[i]->ack, bit);
         __set_bit(src, ppc_cached_irq_mask);
         __pmac_set_irq_mask(src, 0);
@@ -221,7 +221,7 @@ static irqreturn_t gatwick_action(int cp
 			continue;
 		irq += __ilog2(bits);
 		spin_unlock_irqrestore(&pmac_pic_lock, flags);
-		__do_IRQ(irq, irq_desc(irq));
+		__do_IRQ(irq, &irq_desc[irq]);
 		spin_lock_irqsave(&pmac_pic_lock, flags);
 		rc = IRQ_HANDLED;
 	}
Index: linux-2.6/arch/powerpc/platforms/pseries/eeh_driver.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/pseries/eeh_driver.c
+++ linux-2.6/arch/powerpc/platforms/pseries/eeh_driver.c
@@ -70,7 +70,7 @@ static int irq_in_use(unsigned int irq)
 {
 	int rc = 0;
 	unsigned long flags;
-   struct irq_desc *desc = irq_desc(irq);
+   struct irq_desc *desc = irq_desc + irq;
 
 	spin_lock_irqsave(&desc->lock, flags);
 	if (desc->action)
Index: linux-2.6/arch/powerpc/platforms/pseries/xics.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/pseries/xics.c
+++ linux-2.6/arch/powerpc/platforms/pseries/xics.c
@@ -205,7 +205,7 @@ static int get_irq_server(unsigned int v
 {
 	int server;
 	/* For the moment only implement delivery to all cpus or one cpu */
-	cpumask_t cpumask = irq_desc(virq)->affinity;
+	cpumask_t cpumask = irq_desc[virq].affinity;
 	cpumask_t tmp = CPU_MASK_NONE;
 
 	if (! cpu_isset(default_server, cpu_online_map))
@@ -808,7 +808,7 @@ void xics_migrate_irqs_away(void)
 		       virq, cpu);
 
 		/* Reset affinity to all cpus */
-		irq_desc(virq)->affinity = CPU_MASK_ALL;
+		irq_desc[virq].affinity = CPU_MASK_ALL;
 		desc->chip->set_affinity(virq, CPU_MASK_ALL);
 unlock:
 		spin_unlock_irqrestore(&desc->lock, flags);
Index: linux-2.6/arch/powerpc/sysdev/cpm2_pic.c
===================================================================
--- linux-2.6.orig/arch/powerpc/sysdev/cpm2_pic.c
+++ linux-2.6/arch/powerpc/sysdev/cpm2_pic.c
@@ -117,10 +117,9 @@ static void cpm2_end_irq(unsigned int vi
 {
 	int	bit, word;
 	unsigned int irq_nr = virq_to_hw(virq);
-	struct irq_desc *desc = irq_desc(irq_nr);
 
-	if (!(desc->status & (IRQ_DISABLED|IRQ_INPROGRESS))
-			&& desc->action) {
+	if (!(irq_desc[irq_nr].status & (IRQ_DISABLED|IRQ_INPROGRESS))
+			&& irq_desc[irq_nr].action) {
 
 		bit = irq_to_siubit[irq_nr];
 		word = irq_to_siureg[irq_nr];
Index: linux-2.6/arch/powerpc/sysdev/mpic.c
===================================================================
--- linux-2.6.orig/arch/powerpc/sysdev/mpic.c
+++ linux-2.6/arch/powerpc/sysdev/mpic.c
@@ -575,7 +575,7 @@ static struct mpic *mpic_find(unsigned i
 	if (irq < NUM_ISA_INTERRUPTS)
 		return NULL;
 
-	mpic = irq_desc(irq)->chip_data;
+	mpic = irq_desc[irq].chip_data;
 
 	if (is_ipi)
 		*is_ipi = (src >= mpic->ipi_vecs[0] &&
@@ -599,14 +599,14 @@ static inline u32 mpic_physmask(u32 cpum
 /* Get the mpic structure from the IPI number */
 static inline struct mpic * mpic_from_ipi(unsigned int ipi)
 {
-	return irq_desc(ipi)->chip_data;
+	return irq_desc[ipi].chip_data;
 }
 #endif
 
 /* Get the mpic structure from the irq number */
 static inline struct mpic * mpic_from_irq(unsigned int irq)
 {
-	return irq_desc(irq)->chip_data;
+	return irq_desc[irq].chip_data;
 }
 
 /* Send an EOI */
@@ -697,7 +697,7 @@ static void mpic_unmask_ht_irq(unsigned
 
 	mpic_unmask_irq(irq);
 
-	if (irq_desc(irq)->status & IRQ_LEVEL)
+	if (irq_desc[irq].status & IRQ_LEVEL)
 		mpic_ht_end_irq(mpic, src);
 }
 
@@ -707,7 +707,7 @@ static unsigned int mpic_startup_ht_irq(
 	unsigned int src = mpic_irq_to_hw(irq);
 
 	mpic_unmask_irq(irq);
-	mpic_startup_ht_interrupt(mpic, src, irq_desc(irq)->status);
+	mpic_startup_ht_interrupt(mpic, src, irq_desc[irq].status);
 
 	return 0;
 }
@@ -717,7 +717,7 @@ static void mpic_shutdown_ht_irq(unsigne
 	struct mpic *mpic = mpic_from_irq(irq);
 	unsigned int src = mpic_irq_to_hw(irq);
 
-	mpic_shutdown_ht_interrupt(mpic, src, irq_desc(irq)->status);
+	mpic_shutdown_ht_interrupt(mpic, src, irq_desc[irq].status);
 	mpic_mask_irq(irq);
 }
 
@@ -734,7 +734,7 @@ static void mpic_end_ht_irq(unsigned int
 	 * latched another edge interrupt coming in anyway
 	 */
 
-	if (irq_desc(irq)->status & IRQ_LEVEL)
+	if (irq_desc[irq].status & IRQ_LEVEL)
 		mpic_ht_end_irq(mpic, src);
 	mpic_eoi(mpic);
 }
@@ -1368,7 +1368,7 @@ void mpic_setup_this_cpu(void)
  	/* let the mpic know we want intrs. default affinity is 0xffffffff
 	 * until changed via /proc. That's how it's done on x86. If we want
 	 * it differently, then we should make sure we also change the default
-	 * values of irq_desc()->affinity in irq.c.
+	 * values of irq_desc[].affinity in irq.c.
  	 */
 	if (distribute_irqs) {
 	 	for (i = 0; i < mpic->num_sources ; i++)
Index: linux-2.6/arch/powerpc/sysdev/qe_lib/qe_ic.c
===================================================================
--- linux-2.6.orig/arch/powerpc/sysdev/qe_lib/qe_ic.c
+++ linux-2.6/arch/powerpc/sysdev/qe_lib/qe_ic.c
@@ -189,7 +189,7 @@ static inline void qe_ic_write(volatile
 
 static inline struct qe_ic *qe_ic_from_irq(unsigned int virq)
 {
-	return irq_desc(virq)->chip_data;
+	return irq_desc[virq].chip_data;
 }
 
 #define virq_to_hw(virq)	((unsigned int)irq_map[virq].hwirq)
Index: linux-2.6/arch/sh/boards/mach-cayman/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-cayman/irq.c
+++ linux-2.6/arch/sh/boards/mach-cayman/irq.c
@@ -96,7 +96,7 @@ static void ack_cayman_irq(unsigned int
 
 static void end_cayman_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_cayman_irq(irq);
 }
 
@@ -188,7 +188,7 @@ void init_cayman_irq(void)
 	}
 
 	for (i=0; i<NR_EXT_IRQS; i++) {
-		irq_desc(START_EXT_IRQS + i)->chip = &cayman_irq_type;
+		irq_desc[START_EXT_IRQS + i].chip = &cayman_irq_type;
 	}
 
 	/* Setup the SMSC interrupt */
Index: linux-2.6/arch/sh/boards/mach-dreamcast/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-dreamcast/irq.c
+++ linux-2.6/arch/sh/boards/mach-dreamcast/irq.c
@@ -86,7 +86,7 @@ static void ack_systemasic_irq(unsigned
 /* After a IRQ has been ack'd and responded to, it needs to be renabled */
 static void end_systemasic_irq(unsigned int irq)
 {
-        if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+        if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
                 enable_systemasic_irq(irq);
 }
 
Index: linux-2.6/arch/sh/boards/mach-dreamcast/setup.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-dreamcast/setup.c
+++ linux-2.6/arch/sh/boards/mach-dreamcast/setup.c
@@ -47,7 +47,7 @@ static void __init dreamcast_setup(char
 
 	/* Assign all virtual IRQs to the System ASIC int. handler */
 	for (i = HW_EVENT_IRQ_BASE; i < HW_EVENT_IRQ_MAX; i++)
-		irq_desc(i)->chip = &systemasic_int;
+		irq_desc[i].chip = &systemasic_int;
 
 	board_time_init = aica_time_init;
 
Index: linux-2.6/arch/sh/boards/mach-microdev/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-microdev/irq.c
+++ linux-2.6/arch/sh/boards/mach-microdev/irq.c
@@ -134,7 +134,7 @@ static void enable_microdev_irq(unsigned
 static void __init make_microdev_irq(unsigned int irq)
 {
 	disable_irq_nosync(irq);
-	irq_desc(irq)->chip = &microdev_irq_type;
+	irq_desc[irq].chip = &microdev_irq_type;
 	disable_microdev_irq(irq);
 }
 
@@ -145,7 +145,7 @@ static void mask_and_ack_microdev(unsign
 
 static void end_microdev_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_microdev_irq(irq);
 }
 
Index: linux-2.6/arch/sh/boards/mach-se/7206/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-se/7206/irq.c
+++ linux-2.6/arch/sh/boards/mach-se/7206/irq.c
@@ -90,7 +90,7 @@ static void eoi_se7206_irq(unsigned int
 {
 	unsigned short sts0,sts1;
 
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_se7206_irq(irq);
 	/* FPGA isr clear */
 	sts0 = ctrl_inw(INTSTS0);
Index: linux-2.6/arch/sh/boards/mach-se/7343/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-se/7343/irq.c
+++ linux-2.6/arch/sh/boards/mach-se/7343/irq.c
@@ -45,7 +45,7 @@ static void se7343_irq_demux(unsigned in
 
 	while (intv) {
 		if (intv & 1) {
-			ext_desc = irq_desc(ext_irq);
+			ext_desc = irq_desc + ext_irq;
 			handle_level_irq(ext_irq, ext_desc);
 		}
 		intv >>= 1;
Index: linux-2.6/arch/sh/boards/mach-se/7722/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-se/7722/irq.c
+++ linux-2.6/arch/sh/boards/mach-se/7722/irq.c
@@ -45,7 +45,7 @@ static void se7722_irq_demux(unsigned in
 
 	while (intv) {
 		if (intv & 1) {
-			ext_desc = irq_desc(ext_irq);
+			ext_desc = irq_desc + ext_irq;
 			handle_level_irq(ext_irq, ext_desc);
 		}
 		intv >>= 1;
Index: linux-2.6/arch/sh/boards/mach-systemh/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/boards/mach-systemh/irq.c
+++ linux-2.6/arch/sh/boards/mach-systemh/irq.c
@@ -88,14 +88,14 @@ static void mask_and_ack_systemh(unsigne
 
 static void end_systemh_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_systemh_irq(irq);
 }
 
 void make_systemh_irq(unsigned int irq)
 {
 	disable_irq_nosync(irq);
-	irq_desc(irq)->chip = &systemh_irq_type;
+	irq_desc[irq].chip = &systemh_irq_type;
 	disable_systemh_irq(irq);
 }
 
Index: linux-2.6/arch/sh/cchips/hd6446x/hd64461.c
===================================================================
--- linux-2.6.orig/arch/sh/cchips/hd6446x/hd64461.c
+++ linux-2.6/arch/sh/cchips/hd6446x/hd64461.c
@@ -48,7 +48,7 @@ static void mask_and_ack_hd64461(unsigne
 
 static void end_hd64461_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_hd64461_irq(irq);
 }
 
@@ -153,7 +153,7 @@ int __init setup_hd64461(void)
 
 	/*  IRQ 80 -> 95 belongs to HD64461  */
 	for (i = HD64461_IRQBASE; i < HD64461_IRQBASE + 16; i++) {
-		irq_desc(i)->chip = &hd64461_irq_type;
+		irq_desc[i].chip = &hd64461_irq_type;
 	}
 
 	setup_irq(CONFIG_HD64461_IRQ, &irq0);
Index: linux-2.6/arch/sh/cchips/hd6446x/hd64465/setup.c
===================================================================
--- linux-2.6.orig/arch/sh/cchips/hd6446x/hd64465/setup.c
+++ linux-2.6/arch/sh/cchips/hd6446x/hd64465/setup.c
@@ -50,7 +50,7 @@ static void mask_and_ack_hd64465(unsigne
 
 static void end_hd64465_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_hd64465_irq(irq);
 }
 
@@ -164,7 +164,7 @@ static int __init setup_hd64465(void)
 	outw(0xffff, HD64465_REG_NIMR);	/* mask all interrupts */
 
 	for (i = 0; i < HD64465_IRQ_NUM ; i++) {
-		irq_desc(HD64465_IRQ_BASE + i)->chip = &hd64465_irq_type;
+		irq_desc[HD64465_IRQ_BASE + i].chip = &hd64465_irq_type;
 	}
 
 	setup_irq(CONFIG_HD64465_IRQ, &irq0);
Index: linux-2.6/arch/sh/kernel/cpu/irq/imask.c
===================================================================
--- linux-2.6.orig/arch/sh/kernel/cpu/irq/imask.c
+++ linux-2.6/arch/sh/kernel/cpu/irq/imask.c
@@ -96,7 +96,7 @@ static void mask_and_ack_imask(unsigned
 
 static void end_imask_irq(unsigned int irq)
 {
-	if (!(irq_desc(irq)->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
+	if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		enable_imask_irq(irq);
 }
 
@@ -108,6 +108,6 @@ static void shutdown_imask_irq(unsigned
 void make_imask_irq(unsigned int irq)
 {
 	disable_irq_nosync(irq);
-	irq_desc(irq)->chip = &imask_irq_type;
+	irq_desc[irq].chip = &imask_irq_type;
 	enable_irq(irq);
 }
Index: linux-2.6/arch/sh/kernel/cpu/irq/intc-sh5.c
===================================================================
--- linux-2.6.orig/arch/sh/kernel/cpu/irq/intc-sh5.c
+++ linux-2.6/arch/sh/kernel/cpu/irq/intc-sh5.c
@@ -156,7 +156,7 @@ static void end_intc_irq(unsigned int ir
 void make_intc_irq(unsigned int irq)
 {
 	disable_irq_nosync(irq);
-	irq_desc(irq)->chip = &intc_irq_type;
+	irq_desc[irq].chip = &intc_irq_type;
 	disable_intc_irq(irq);
 }
 
@@ -196,7 +196,7 @@ void __init plat_irq_setup(void)
 
 	/* Set default: per-line enable/disable, priority driven ack/eoi */
 	for (i = 0; i < NR_INTC_IRQS; i++)
-		irq_desc(i)->chip = &intc_irq_type;
+		irq_desc[i].chip = &intc_irq_type;
 
 
 	/* Disable all interrupts and set all priorities to 0 to avoid trouble */
Index: linux-2.6/arch/sh/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/kernel/irq.c
+++ linux-2.6/arch/sh/kernel/irq.c
@@ -45,24 +45,22 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < sh_mv.mv_nr_irqs) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 		seq_printf(p, "%3d: ",i);
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
-		seq_printf(p, " %14s", desc->chip->name);
-		seq_printf(p, "-%-8s", desc->name);
+		seq_printf(p, " %14s", irq_desc[i].chip->name);
+		seq_printf(p, "-%-8s", irq_desc[i].name);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
 		seq_putc(p, '\n');
 unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == sh_mv.mv_nr_irqs)
 		seq_printf(p, "Err: %10u\n", atomic_read(&irq_err_count));
 
Index: linux-2.6/arch/sparc64/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/sparc64/kernel/irq.c
+++ linux-2.6/arch/sparc64/kernel/irq.c
@@ -176,10 +176,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ",i);
@@ -189,7 +187,7 @@ int show_interrupts(struct seq_file *p,
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
-		seq_printf(p, " %9s", desc->chip->typename);
+		seq_printf(p, " %9s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -197,7 +195,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	}
 	return 0;
 }
@@ -249,7 +247,7 @@ struct irq_handler_data {
 #ifdef CONFIG_SMP
 static int irq_choose_cpu(unsigned int virt_irq)
 {
-	cpumask_t mask = irq_desc(virt_irq)->affinity;
+	cpumask_t mask = irq_desc[virt_irq].affinity;
 	int cpuid;
 
 	if (cpus_equal(mask, CPU_MASK_ALL)) {
@@ -335,7 +333,7 @@ static void sun4u_irq_disable(unsigned i
 static void sun4u_irq_eoi(unsigned int virt_irq)
 {
 	struct irq_handler_data *data = get_irq_chip_data(virt_irq);
-	struct irq_desc *desc = irq_desc(virt_irq);
+	struct irq_desc *desc = irq_desc + virt_irq;
 
 	if (unlikely(desc->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
 		return;
@@ -390,7 +388,7 @@ static void sun4v_irq_disable(unsigned i
 static void sun4v_irq_eoi(unsigned int virt_irq)
 {
 	unsigned int ino = virt_irq_table[virt_irq].dev_ino;
-	struct irq_desc *desc = irq_desc(virt_irq);
+	struct irq_desc *desc = irq_desc + virt_irq;
 	int err;
 
 	if (unlikely(desc->status & (IRQ_DISABLED|IRQ_INPROGRESS)))
@@ -466,7 +464,7 @@ static void sun4v_virq_disable(unsigned
 
 static void sun4v_virq_eoi(unsigned int virt_irq)
 {
-	struct irq_desc *desc = irq_desc(virt_irq);
+	struct irq_desc *desc = irq_desc + virt_irq;
 	unsigned long dev_handle, dev_ino;
 	int err;
 
@@ -524,7 +522,7 @@ void irq_install_pre_handler(int virt_ir
 			     void *arg1, void *arg2)
 {
 	struct irq_handler_data *data = get_irq_chip_data(virt_irq);
-	struct irq_desc *desc = irq_desc(virt_irq);
+	struct irq_desc *desc = irq_desc + virt_irq;
 
 	data->pre_handler = func;
 	data->arg1 = arg1;
@@ -650,7 +648,7 @@ unsigned int sun4v_build_virq(u32 devhan
 	 * especially wrt. locking, we do not let request_irq() enable
 	 * the interrupt.
 	 */
-	desc = irq_desc(virt_irq);
+	desc = irq_desc + virt_irq;
 	desc->status |= IRQ_NOAUTOEN;
 
 	set_irq_chip_data(virt_irq, data);
@@ -738,7 +736,7 @@ void handler_irq(int irq, struct pt_regs
 		virt_irq = bucket_get_virt_irq(bucket_pa);
 		bucket_clear_chain_pa(bucket_pa);
 
-		desc = irq_desc(virt_irq);
+		desc = irq_desc + virt_irq;
 
 		desc->handle_irq(virt_irq, desc);
 
@@ -784,16 +782,15 @@ void fixup_irqs(void)
 
 	for (irq = 0; irq < NR_IRQS; irq++) {
 		unsigned long flags;
-		struct irq_desc *desc = irq_desc(irq);
 
-		spin_lock_irqsave(&desc->lock, flags);
-		if (desc->action &&
-		    !(desc->status & IRQ_PER_CPU)) {
-			if (desc->chip->set_affinity)
-				desc->chip->set_affinity(irq,
-					desc->affinity);
+		spin_lock_irqsave(&irq_desc[irq].lock, flags);
+		if (irq_desc[irq].action &&
+		    !(irq_desc[irq].status & IRQ_PER_CPU)) {
+			if (irq_desc[irq].chip->set_affinity)
+				irq_desc[irq].chip->set_affinity(irq,
+					irq_desc[irq].affinity);
 		}
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[irq].lock, flags);
 	}
 }
 #endif
@@ -1035,5 +1032,5 @@ void __init init_IRQ(void)
 			     : "i" (PSTATE_IE)
 			     : "g1");
 
-	irq_desc(0)->action = &timer_irq_action;
+	irq_desc[0].action = &timer_irq_action;
 }
Index: linux-2.6/arch/sparc64/kernel/pci_msi.c
===================================================================
--- linux-2.6.orig/arch/sparc64/kernel/pci_msi.c
+++ linux-2.6/arch/sparc64/kernel/pci_msi.c
@@ -33,7 +33,7 @@ static irqreturn_t sparc64_msiq_interrup
 			unsigned int virt_irq;
 
 			virt_irq = pbm->msi_irq_table[msi - pbm->msi_first];
-			desc = irq_desc(virt_irq);
+			desc = irq_desc + virt_irq;
 
 			desc->handle_irq(virt_irq, desc);
 		}
Index: linux-2.6/arch/um/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/um/kernel/irq.c
+++ linux-2.6/arch/um/kernel/irq.c
@@ -33,10 +33,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ",i);
@@ -46,7 +44,7 @@ int show_interrupts(struct seq_file *p,
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
-		seq_printf(p, " %14s", desc->chip->typename);
+		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -54,7 +52,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS)
 		seq_putc(p, '\n');
 
@@ -334,7 +332,7 @@ unsigned int do_IRQ(int irq, struct uml_
 {
 	struct pt_regs *old_regs = set_irq_regs((struct pt_regs *)regs);
 	irq_enter();
-	__do_IRQ(irq, irq_desc(irq));
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 	set_irq_regs(old_regs);
 	return 1;
@@ -390,20 +388,17 @@ static struct hw_interrupt_type SIGVTALR
 void __init init_IRQ(void)
 {
 	int i;
-	struct irq_desc *desc;
 
-	desc  = irq_desc(TIMER_IRQ);
-	desc->status = IRQ_DISABLED;
-	desc->action = NULL;
-	desc->depth = 1;
-	desc->chip = &SIGVTALRM_irq_type;
+	irq_desc[TIMER_IRQ].status = IRQ_DISABLED;
+	irq_desc[TIMER_IRQ].action = NULL;
+	irq_desc[TIMER_IRQ].depth = 1;
+	irq_desc[TIMER_IRQ].chip = &SIGVTALRM_irq_type;
 	enable_irq(TIMER_IRQ);
 	for (i = 1; i < NR_IRQS; i++) {
-		desc = irq_desc(i);
-		desc->status = IRQ_DISABLED;
-		desc->action = NULL;
-		desc->depth = 1;
-		desc->chip = &normal_irq_type;
+		irq_desc[i].status = IRQ_DISABLED;
+		irq_desc[i].action = NULL;
+		irq_desc[i].depth = 1;
+		irq_desc[i].chip = &normal_irq_type;
 		enable_irq(i);
 	}
 }
Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic.c
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -557,7 +557,7 @@ static void set_ioapic_affinity_irq(unsi
 	 */
 	dest = SET_APIC_LOGICAL_ID(dest);
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	spin_lock_irqsave(&ioapic_lock, flags);
 	__target_IO_APIC_irq(irq, dest, cfg->vector);
 	desc->affinity = mask;
@@ -1376,9 +1376,9 @@ static void ioapic_register_intr(int irq
 
 	/* first time to use this irq_desc */
 	if (irq < 16)
-		desc = irq_desc(irq);
+		desc = to_irq_desc(irq);
 	else
-		desc = irq_desc_with_new(irq);
+		desc = to_irq_desc_with_new(irq);
 
 	if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
 	    trigger == IOAPIC_LEVEL)
@@ -2256,7 +2256,7 @@ static void migrate_ioapic_irq(int irq,
 	cpus_and(tmp, cfg->domain, mask);
 	dest = cpu_mask_to_apicid(tmp);
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	modify_ioapic_rte = desc->status & IRQ_LEVEL;
 	if (modify_ioapic_rte) {
 		spin_lock_irqsave(&ioapic_lock, flags);
@@ -2285,7 +2285,7 @@ static void migrate_ioapic_irq(int irq,
 static int migrate_irq_remapped_level(int irq)
 {
 	int ret = -1;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	mask_IO_APIC_irq(irq);
 
@@ -2340,7 +2340,7 @@ static void ir_irq_migration(struct work
  */
 static void set_ir_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	if (desc->status & IRQ_LEVEL) {
 		desc->status |= IRQ_MOVE_PENDING;
@@ -2369,7 +2369,7 @@ asmlinkage void smp_irq_move_cleanup_int
 		struct irq_cfg *cfg;
 		irq = __get_cpu_var(vector_irq)[vector];
 
-		desc = irq_desc(irq);
+		desc = to_irq_desc(irq);
 		if (!desc)
 			continue;
 
@@ -2426,7 +2426,7 @@ static void ack_x2apic_edge(unsigned int
 
 static void ack_apic_edge(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	irq_complete_move(irq);
 	move_native_irq(irq, desc);
@@ -2444,7 +2444,7 @@ static void ack_apic_level(unsigned int
 	int i;
 #endif
 	int do_unmask_irq = 0;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	irq_complete_move(irq);
 #ifdef CONFIG_GENERIC_PENDING_IRQ
@@ -2585,7 +2585,7 @@ static inline void init_IO_APIC_traps(vo
 			if (irq < 16)
 				make_8259A_irq(irq);
 			else {
-				desc = irq_desc(irq);
+				desc = to_irq_desc(irq);
 				/* Strange. Oh, well.. */
 				desc->chip = &no_irq_chip;
 			}
@@ -2629,7 +2629,7 @@ static void lapic_register_intr(int irq)
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->status &= ~IRQ_LEVEL;
 	set_irq_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq,
 				      "edge");
@@ -3189,7 +3189,7 @@ static void set_msi_irq_affinity(unsigne
 	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
 
 	write_msi_msg(irq, &msg);
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->affinity = mask;
 }
 
@@ -3240,7 +3240,7 @@ static void ir_set_msi_irq_affinity(unsi
 		cfg->move_in_progress = 0;
 	}
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->affinity = mask;
 }
 #endif
@@ -3315,7 +3315,7 @@ static int setup_msi_irq(struct pci_dev
 
 #ifdef CONFIG_INTR_REMAP
 	if (irq_remapped(irq)) {
-		struct irq_desc *desc = irq_desc(irq);
+		struct irq_desc *desc = to_irq_desc(irq);
 		/*
 		 * irq migration in process context
 		 */
@@ -3469,7 +3469,7 @@ static void dmar_msi_set_affinity(unsign
 
 	dmar_msi_write(irq, &msg);
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->affinity = mask;
 }
 #endif /* CONFIG_SMP */
@@ -3541,7 +3541,7 @@ static void set_ht_irq_affinity(unsigned
 	dest = cpu_mask_to_apicid(tmp);
 
 	target_ht_irq(irq, dest, cfg->vector);
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->affinity = mask;
 }
 #endif
Index: linux-2.6/arch/x86/kernel/irq_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irq_32.c
+++ linux-2.6/arch/x86/kernel/irq_32.c
@@ -235,7 +235,7 @@ unsigned int do_IRQ(struct pt_regs *regs
 
 	overflow = check_stack_overflow();
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (unlikely(!desc)) {
 		printk(KERN_EMERG "%s: cannot handle IRQ %d vector %#x cpu %d\n",
 					__func__, irq, vector, smp_processor_id());
@@ -284,7 +284,7 @@ int show_interrupts(struct seq_file *p,
 	if (i == nr_irqs)
 		tail = 1;
 	else
-		desc = irq_desc(i);
+		desc = to_irq_desc(i);
 #endif
 
 	if (i == 0) {
Index: linux-2.6/arch/x86/kernel/irq_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irq_64.c
+++ linux-2.6/arch/x86/kernel/irq_64.c
@@ -89,7 +89,7 @@ int show_interrupts(struct seq_file *p,
 	if (i == nr_irqs)
 		tail = 1;
 	else
-		desc = irq_desc(i);
+		desc = to_irq_desc(i);
 #endif
 
 	if (i == 0) {
@@ -223,7 +223,7 @@ asmlinkage unsigned int do_IRQ(struct pt
 	stack_overflow_check(regs);
 #endif
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (likely(desc))
 		generic_handle_irq_desc(irq, desc);
 	else {
Index: linux-2.6/arch/x86/kernel/irqinit_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irqinit_32.c
+++ linux-2.6/arch/x86/kernel/irqinit_32.c
@@ -70,7 +70,7 @@ void __init init_ISA_irqs (void)
 	 */
 	for (i = 0; i < 16; i++) {
 		/* first time call this irq_desc */
-		struct irq_desc *desc = irq_desc_with_new(i);
+		struct irq_desc *desc = to_irq_desc_with_new(i);
 
 		desc->status = IRQ_DISABLED;
 		desc->action = NULL;
Index: linux-2.6/arch/x86/kernel/irqinit_64.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/irqinit_64.c
+++ linux-2.6/arch/x86/kernel/irqinit_64.c
@@ -144,7 +144,7 @@ static void __init init_ISA_irqs (void)
 
 	for (i = 0; i < 16; i++) {
 		/* first time call this irq_desc */
-		struct irq_desc *desc = irq_desc_with_new(i);
+		struct irq_desc *desc = to_irq_desc_with_new(i);
 
 		desc->status = IRQ_DISABLED;
 		desc->action = NULL;
Index: linux-2.6/arch/x86/kernel/visws_quirks.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/visws_quirks.c
+++ linux-2.6/arch/x86/kernel/visws_quirks.c
@@ -484,7 +484,7 @@ static void disable_cobalt_irq(unsigned
 static unsigned int startup_cobalt_irq(unsigned int irq)
 {
 	unsigned long flags;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	spin_lock_irqsave(&cobalt_lock, flags);
 	if ((desc->status & (IRQ_DISABLED | IRQ_INPROGRESS | IRQ_WAITING)))
@@ -507,7 +507,7 @@ static void ack_cobalt_irq(unsigned int
 static void end_cobalt_irq(unsigned int irq)
 {
 	unsigned long flags;
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	spin_lock_irqsave(&cobalt_lock, flags);
 	if (!(desc->status & (IRQ_DISABLED | IRQ_INPROGRESS)))
@@ -628,7 +628,7 @@ static irqreturn_t piix4_master_intr(int
 
 	spin_unlock_irqrestore(&i8259A_lock, flags);
 
-	desc = irq_desc(realirq);
+	desc = to_irq_desc(realirq);
 
 	/*
 	 * handle this 'virtual interrupt' as a Cobalt one now.
@@ -664,7 +664,7 @@ void init_VISWS_APIC_irqs(void)
 	int i;
 
 	for (i = 0; i < CO_IRQ_APIC0 + CO_APIC_LAST + 1; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = to_irq_desc(i);
 
 		desc->status = IRQ_DISABLED;
 		desc->action = 0;
Index: linux-2.6/arch/x86/mach-voyager/voyager_smp.c
===================================================================
--- linux-2.6.orig/arch/x86/mach-voyager/voyager_smp.c
+++ linux-2.6/arch/x86/mach-voyager/voyager_smp.c
@@ -1481,7 +1481,7 @@ static void disable_local_vic_irq(unsign
  * the interrupt off to another CPU */
 static void before_handle_vic_irq(unsigned int irq)
 {
-	irq_desc_t *desc = irq_desc(irq);
+	irq_desc_t *desc = to_irq_desc(irq);
 	__u8 cpu = smp_processor_id();
 
 	_raw_spin_lock(&vic_irq_lock);
@@ -1516,7 +1516,7 @@ static void before_handle_vic_irq(unsign
 /* Finish the VIC interrupt: basically mask */
 static void after_handle_vic_irq(unsigned int irq)
 {
-	irq_desc_t *desc = irq_desc(irq);
+	irq_desc_t *desc = to_irq_desc(irq);
 
 	_raw_spin_lock(&vic_irq_lock);
 	{
Index: linux-2.6/arch/xtensa/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/xtensa/kernel/irq.c
+++ linux-2.6/arch/xtensa/kernel/irq.c
@@ -44,7 +44,7 @@ void ack_bad_irq(unsigned int irq)
 asmlinkage void do_IRQ(int irq, struct pt_regs *regs)
 {
 	struct pt_regs *old_regs = set_irq_regs(regs);
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = irq_desc + irq;
 
 	if (irq >= NR_IRQS) {
 		printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
@@ -90,10 +90,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto skip;
 		seq_printf(p, "%3d: ",i);
@@ -103,7 +101,7 @@ int show_interrupts(struct seq_file *p,
 		for_each_online_cpu(j)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
-		seq_printf(p, " %14s", desc->chip->typename);
+		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
 
 		for (action=action->next; action; action = action->next)
@@ -111,7 +109,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 skip:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 		seq_printf(p, "NMI: ");
 		for_each_online_cpu(j)
Index: linux-2.6/drivers/char/random.c
===================================================================
--- linux-2.6.orig/drivers/char/random.c
+++ linux-2.6/drivers/char/random.c
@@ -588,7 +588,7 @@ static struct timer_rand_state *get_time
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 
 	if (!desc)
 		return NULL;
@@ -600,7 +600,7 @@ static void set_timer_rand_state(unsigne
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 
 	if (!desc)
 		return;
Index: linux-2.6/drivers/gpio/gpiolib.c
===================================================================
--- linux-2.6.orig/drivers/gpio/gpiolib.c
+++ linux-2.6/drivers/gpio/gpiolib.c
@@ -1058,7 +1058,7 @@ static void gpiolib_dbg_show(struct seq_
 
 		if (!is_out) {
 			int		irq = gpio_to_irq(gpio);
-			struct irq_desc	*desc = irq_desc(irq);
+			struct irq_desc	*desc = to_irq_desc(irq);
 
 			/* This races with request_irq(), set_irq_type(),
 			 * and set_irq_wake() ... but those are "rare".
Index: linux-2.6/drivers/mfd/asic3.c
===================================================================
--- linux-2.6.orig/drivers/mfd/asic3.c
+++ linux-2.6/drivers/mfd/asic3.c
@@ -124,7 +124,7 @@ static void asic3_irq_demux(unsigned int
 					irqnr = asic->irq_base +
 						(ASIC3_GPIOS_PER_BANK * bank)
 						+ i;
-					desc = irq_desc(irqnr);
+					desc = to_irq_desc(irqnr);
 					desc->handle_irq(irqnr, desc);
 					if (asic->irq_bothedge[bank] & bit)
 						asic3_irq_flip_edge(asic, base,
@@ -137,7 +137,7 @@ static void asic3_irq_demux(unsigned int
 		for (i = ASIC3_NUM_GPIOS; i < ASIC3_NR_IRQS; i++) {
 			/* They start at bit 4 and go up */
 			if (status & (1 << (i - ASIC3_NUM_GPIOS + 4))) {
-				desc = irq_desc(asic->irq_base + i);
+				desc = to_irq_desc(asic->irq_base + i);
 				desc->handle_irq(asic->irq_base + i,
 						 desc);
 			}
Index: linux-2.6/drivers/mfd/htc-egpio.c
===================================================================
--- linux-2.6.orig/drivers/mfd/htc-egpio.c
+++ linux-2.6/drivers/mfd/htc-egpio.c
@@ -112,7 +112,7 @@ static void egpio_handler(unsigned int i
 		/* Run irq handler */
 		pr_debug("got IRQ %d\n", irqpin);
 		irq = ei->irq_start + irqpin;
-		desc = irq_desc(irq);
+		desc = to_irq_desc(irq);
 		desc->handle_irq(irq, desc);
 	}
 }
Index: linux-2.6/drivers/parisc/dino.c
===================================================================
--- linux-2.6.orig/drivers/parisc/dino.c
+++ linux-2.6/drivers/parisc/dino.c
@@ -298,7 +298,7 @@ struct pci_port_ops dino_port_ops = {
 
 static void dino_disable_irq(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	struct dino_device *dino_dev = desc->chip_data;
 	int local_irq = gsc_find_local_irq(irq, dino_dev->global_irq, DINO_LOCAL_IRQS);
 
@@ -311,7 +311,7 @@ static void dino_disable_irq(unsigned in
 
 static void dino_enable_irq(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	struct dino_device *dino_dev = desc->chip_data;
 	int local_irq = gsc_find_local_irq(irq, dino_dev->global_irq, DINO_LOCAL_IRQS);
 	u32 tmp;
@@ -389,7 +389,7 @@ ilr_again:
 	do {
 		int local_irq = __ffs(mask);
 		int irq = dino_dev->global_irq[local_irq];
-		struct irq_desc *desc = irq_desc(irq);
+		struct irq_desc *desc = to_irq_desc(irq);
 		DBG(KERN_DEBUG "%s(%d, %p) mask 0x%x\n",
 			__func__, irq, intr_dev, mask);
 		__do_IRQ(irq, desc);
Index: linux-2.6/drivers/parisc/eisa.c
===================================================================
--- linux-2.6.orig/drivers/parisc/eisa.c
+++ linux-2.6/drivers/parisc/eisa.c
@@ -234,7 +234,7 @@ static irqreturn_t eisa_irq(int wax_irq,
 	}
 	spin_unlock_irqrestore(&eisa_irq_lock, flags);
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	__do_IRQ(irq, desc);
    
 	spin_lock_irqsave(&eisa_irq_lock, flags);
@@ -348,10 +348,10 @@ static int __init eisa_probe(struct pari
 	}
 	
 	/* Reserve IRQ2 */
-	irq_desc(2)->action = &irq2_action;
+	to_irq_desc(2)->action = &irq2_action;
 	
 	for (i = 0; i < 16; i++) {
-		irq_desc(i)->chip = &eisa_interrupt_type;
+		to_irq_desc(i)->chip = &eisa_interrupt_type;
 	}
 	
 	EISA_bus = 1;
Index: linux-2.6/drivers/parisc/gsc.c
===================================================================
--- linux-2.6.orig/drivers/parisc/gsc.c
+++ linux-2.6/drivers/parisc/gsc.c
@@ -87,7 +87,7 @@ irqreturn_t gsc_asic_intr(int gsc_asic_i
 	do {
 		int local_irq = __ffs(irr);
 		unsigned int irq = gsc_asic->global_irq[local_irq];
-		struct irq_desc *desc = irq_desc(irq);
+		struct irq_desc *desc = to_irq_desc(irq);
 		__do_IRQ(irq, desc);
 		irr &= ~(1 << local_irq);
 	} while (irr);
@@ -109,7 +109,7 @@ int gsc_find_local_irq(unsigned int irq,
 
 static void gsc_asic_disable_irq(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	struct gsc_asic *irq_dev = desc->chip_data;
 	int local_irq = gsc_find_local_irq(irq, irq_dev->global_irq, 32);
 	u32 imr;
@@ -125,7 +125,7 @@ static void gsc_asic_disable_irq(unsigne
 
 static void gsc_asic_enable_irq(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	struct gsc_asic *irq_dev = desc->chip_data;
 	int local_irq = gsc_find_local_irq(irq, irq_dev->global_irq, 32);
 	u32 imr;
@@ -167,7 +167,7 @@ int gsc_assign_irq(struct hw_interrupt_t
 	if (irq > GSC_IRQ_MAX)
 		return NO_IRQ;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->chip = type;
 	desc->chip_data = data;
 	return irq++;
Index: linux-2.6/drivers/parisc/iosapic.c
===================================================================
--- linux-2.6.orig/drivers/parisc/iosapic.c
+++ linux-2.6/drivers/parisc/iosapic.c
@@ -619,7 +619,7 @@ iosapic_set_irt_data( struct vector_info
 
 static struct vector_info *iosapic_get_vector(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	return desc->chip_data;
 }
Index: linux-2.6/drivers/parisc/superio.c
===================================================================
--- linux-2.6.orig/drivers/parisc/superio.c
+++ linux-2.6/drivers/parisc/superio.c
@@ -140,7 +140,7 @@ superio_interrupt(int parent_irq, void *
 	}
 
 	/* Call the appropriate device's interrupt */
-	desc = irq_desc(local_irq);
+	desc = to_irq_desc(local_irq);
 	__do_IRQ(local_irq, desc);
 
 	/* set EOI - forces a new interrupt if a lower priority device
@@ -365,7 +365,7 @@ int superio_fixup_irq(struct pci_dev *pc
 #endif
 
 	for (i = 0; i < 16; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = to_irq_desc(i);
 
 		desc->chip = &superio_interrupt_type;
 	}
Index: linux-2.6/drivers/pci/intr_remapping.c
===================================================================
--- linux-2.6.orig/drivers/pci/intr_remapping.c
+++ linux-2.6/drivers/pci/intr_remapping.c
@@ -64,7 +64,7 @@ static struct irq_2_iommu *irq_2_iommu(u
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 
 	BUG_ON(!desc);
 
@@ -76,7 +76,7 @@ static struct irq_2_iommu *irq_2_iommu_w
 	struct irq_desc *desc;
 	struct irq_2_iommu *irq_iommu;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 
 	BUG_ON(!desc);
 
Index: linux-2.6/drivers/pcmcia/hd64465_ss.c
===================================================================
--- linux-2.6.orig/drivers/pcmcia/hd64465_ss.c
+++ linux-2.6/drivers/pcmcia/hd64465_ss.c
@@ -241,7 +241,7 @@ static void hs_map_irq(hs_socket_t *sp,
 	if (irq >= HS_NUM_MAPPED_IRQS)
 	    return;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
     	hs_mapped_irq[irq].sock = sp;
 	/* insert ourselves as the irq controller */
 	hs_mapped_irq[irq].old_handler = desc->chip;
@@ -261,7 +261,7 @@ static void hs_unmap_irq(hs_socket_t *sp
 	if (irq >= HS_NUM_MAPPED_IRQS)
 	    return;
 		
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	/* restore the original irq controller */
 	desc->chip = hs_mapped_irq[irq].old_handler;
 }
Index: linux-2.6/drivers/xen/events.c
===================================================================
--- linux-2.6.orig/drivers/xen/events.c
+++ linux-2.6/drivers/xen/events.c
@@ -125,7 +125,7 @@ static void bind_evtchn_to_cpu(unsigned
 
 	BUG_ON(irq == -1);
 #ifdef CONFIG_SMP
-	irq_desc(irq)->affinity = cpumask_of_cpu(cpu);
+	to_irq_desc(irq)->affinity = cpumask_of_cpu(cpu);
 #endif
 
 	__clear_bit(chn, cpu_evtchn_mask[cpu_evtchn[chn]]);
@@ -140,7 +140,7 @@ static void init_evtchn_cpu_bindings(voi
 	int i;
 	/* By default all event channels notify CPU#0. */
 	for (i = 0; i < nr_irqs; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = to_irq_desc(i);
 		desc->affinity = cpumask_of_cpu(0);
 	}
 #endif
@@ -640,7 +640,7 @@ static void ack_dynirq(unsigned int irq)
 {
 	int evtchn = evtchn_from_irq(irq);
 
-	move_native_irq(irq, irq_desc(irq));
+	move_native_irq(irq, to_irq_desc(irq));
 
 	if (VALID_EVTCHN(evtchn))
 		clear_evtchn(evtchn);
Index: linux-2.6/include/asm-mips/irq.h
===================================================================
--- linux-2.6.orig/include/asm-mips/irq.h
+++ linux-2.6/include/asm-mips/irq.h
@@ -65,7 +65,7 @@ extern void smtc_forward_irq(unsigned in
  */
 #define IRQ_AFFINITY_HOOK(irq)						\
 do {									\
-    if (!cpu_isset(smp_processor_id(), irq_desc(irq)->affinity)) {	\
+    if (!cpu_isset(smp_processor_id(), irq_desc[irq].affinity)) {	\
 	smtc_forward_irq(irq);						\
 	irq_exit();							\
 	return;								\
Index: linux-2.6/include/linux/irq.h
===================================================================
--- linux-2.6.orig/include/linux/irq.h
+++ linux-2.6/include/linux/irq.h
@@ -194,21 +194,21 @@ struct irq_desc {
 	const char		*name;
 } ____cacheline_internodealigned_in_smp;
 
-extern struct irq_desc *irq_desc(unsigned int irq);
-extern struct irq_desc *irq_desc_with_new(unsigned int irq);
+extern struct irq_desc *to_irq_desc(unsigned int irq);
+extern struct irq_desc *to_irq_desc_with_new(unsigned int irq);
 
 #ifndef CONFIG_HAVE_SPARSE_IRQ
 
 #ifndef CONFIG_HAVE_DYN_ARRAY
 /* could be removed if we get rid of all irq_desc reference */
-extern struct irq_desc irq_descX[NR_IRQS];
+extern struct irq_desc irq_desc[NR_IRQS];
 #else
-extern struct irq_desc *irq_descX;
+extern struct irq_desc *irq_desc;
 #endif
 
 #ifdef CONFIG_GENERIC_HARDIRQS
 #define for_each_irq_desc(irq, desc)		\
-	for (irq = 0, desc = irq_descX; irq < nr_irqs; irq++, desc = &irq_descX[irq])
+	for (irq = 0, desc = irq_desc; irq < nr_irqs; irq++, desc = &irq_desc[irq])
 #endif
 
 #else
@@ -280,7 +280,7 @@ static inline int irq_balancing_disabled
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	return desc->status & IRQ_NO_BALANCING_MASK;
 }
 
@@ -325,7 +325,7 @@ static inline void generic_handle_irq_de
 
 static inline void generic_handle_irq(unsigned int irq)
 {
-	generic_handle_irq_desc(irq, irq_desc(irq));
+	generic_handle_irq_desc(irq, to_irq_desc(irq));
 }
 
 /* Handling of unhandled and spurious interrupts: */
@@ -362,7 +362,7 @@ static inline void __set_irq_handler_unl
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->handle_irq = handler;
 }
 
@@ -397,7 +397,7 @@ extern void destroy_irq(unsigned int irq
 /* Test to see if a driver has successfully requested an irq */
 static inline int irq_has_action(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	return desc->action != NULL;
 }
 
@@ -412,10 +412,10 @@ extern int set_irq_chip_data(unsigned in
 extern int set_irq_type(unsigned int irq, unsigned int type);
 extern int set_irq_msi(unsigned int irq, struct msi_desc *entry);
 
-#define get_irq_chip(irq)	(irq_desc(irq)->chip)
-#define get_irq_chip_data(irq)	(irq_desc(irq)->chip_data)
-#define get_irq_data(irq)	(irq_desc(irq)->handler_data)
-#define get_irq_msi(irq)	(irq_desc(irq)->msi_desc)
+#define get_irq_chip(irq)	(to_irq_desc(irq)->chip)
+#define get_irq_chip_data(irq)	(to_irq_desc(irq)->chip_data)
+#define get_irq_data(irq)	(to_irq_desc(irq)->handler_data)
+#define get_irq_msi(irq)	(to_irq_desc(irq)->msi_desc)
 
 #endif /* CONFIG_GENERIC_HARDIRQS */
 
Index: linux-2.6/kernel/irq/autoprobe.c
===================================================================
--- linux-2.6.orig/kernel/irq/autoprobe.c
+++ linux-2.6/kernel/irq/autoprobe.c
@@ -39,7 +39,7 @@ unsigned long probe_irq_on(void)
 	 * flush such a longstanding irq before considering it as spurious.
 	 */
 	for (i = nr_irqs-1; i > 0; i--) {
-		desc = irq_desc(i);
+		desc = to_irq_desc(i);
 		if (!desc)
 			continue;
 
@@ -71,7 +71,7 @@ unsigned long probe_irq_on(void)
 	 * happened in the previous stage, it may have masked itself)
 	 */
 	for (i = nr_irqs-1; i > 0; i--) {
-		desc = irq_desc(i);
+		desc = to_irq_desc(i);
 		if (!desc)
 			continue;
 
@@ -96,7 +96,7 @@ unsigned long probe_irq_on(void)
 	for (i = 0; i < nr_irqs; i++) {
 		unsigned int status;
 
-		desc = irq_desc(i);
+		desc = to_irq_desc(i);
 		if (!desc)
 			continue;
 		spin_lock_irq(&desc->lock);
@@ -137,7 +137,7 @@ unsigned int probe_irq_mask(unsigned lon
 
 	mask = 0;
 	for (i = 0; i < nr_irqs; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = to_irq_desc(i);
 		unsigned int status;
 
 		if (!desc)
@@ -182,7 +182,7 @@ int probe_irq_off(unsigned long val)
 	int i, irq_found = 0, nr_irqs = 0;
 
 	for (i = 0; i < nr_irqs; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = to_irq_desc(i);
 		unsigned int status;
 
 		if (!desc)
Index: linux-2.6/kernel/irq/chip.c
===================================================================
--- linux-2.6.orig/kernel/irq/chip.c
+++ linux-2.6/kernel/irq/chip.c
@@ -28,7 +28,7 @@ void dynamic_irq_init(unsigned int irq)
 	unsigned long flags;
 
 	/* first time to use this irq_desc */
-	desc = irq_desc_with_new(irq);
+	desc = to_irq_desc_with_new(irq);
 	if (!desc) {
 		WARN(1, KERN_ERR "Trying to initialize invalid IRQ%d\n", irq);
 		return;
@@ -61,7 +61,7 @@ void dynamic_irq_cleanup(unsigned int ir
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		WARN(1, KERN_ERR "Trying to cleanup invalid IRQ%d\n", irq);
 		return;
@@ -93,7 +93,7 @@ int set_irq_chip(unsigned int irq, struc
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		WARN(1, KERN_ERR "Trying to install chip for IRQ%d\n", irq);
 		return -EINVAL;
@@ -102,7 +102,7 @@ int set_irq_chip(unsigned int irq, struc
 	if (!chip)
 		chip = &no_irq_chip;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	spin_lock_irqsave(&desc->lock, flags);
 	irq_chip_set_defaults(chip);
 	desc->chip = chip;
@@ -123,7 +123,7 @@ int set_irq_type(unsigned int irq, unsig
 	unsigned long flags;
 	int ret = -ENXIO;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR "Trying to set irq type for IRQ%d\n", irq);
 		return -ENODEV;
@@ -150,7 +150,7 @@ int set_irq_data(unsigned int irq, void
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR
 		       "Trying to install controller data for IRQ%d\n", irq);
@@ -176,7 +176,7 @@ int set_irq_msi(unsigned int irq, struct
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR
 		       "Trying to install msi data for IRQ%d\n", irq);
@@ -203,7 +203,7 @@ int set_irq_chip_data(unsigned int irq,
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR
 		       "Trying to install chip data for IRQ%d\n", irq);
@@ -230,7 +230,7 @@ static void default_enable(unsigned int
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->chip->unmask(irq);
 	desc->status &= ~IRQ_MASKED;
 }
@@ -249,7 +249,7 @@ static unsigned int default_startup(unsi
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->chip->enable(irq);
 
 	return 0;
@@ -262,7 +262,7 @@ static void default_shutdown(unsigned in
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->chip->mask(irq);
 	desc->status |= IRQ_MASKED;
 }
@@ -554,7 +554,7 @@ __set_irq_handler(unsigned int irq, irq_
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR
 		       "Trying to install type control for IRQ%d\n", irq);
@@ -618,7 +618,7 @@ void __init set_irq_noprobe(unsigned int
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR "Trying to mark IRQ%d non-probeable\n", irq);
 
@@ -635,7 +635,7 @@ void __init set_irq_probe(unsigned int i
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc) {
 		printk(KERN_ERR "Trying to mark IRQ%d probeable\n", irq);
 
Index: linux-2.6/kernel/irq/handle.c
===================================================================
--- linux-2.6.orig/kernel/irq/handle.c
+++ linux-2.6/kernel/irq/handle.c
@@ -153,7 +153,7 @@ early_param("nr_irq_desc", parse_nr_irq_
 
 DEFINE_DYN_ARRAY(irq_descX, sizeof(struct irq_desc), nr_irq_desc, PAGE_SIZE, init_work);
 
-struct irq_desc *irq_desc(unsigned int irq)
+struct irq_desc *to_irq_desc(unsigned int irq)
 {
 	struct irq_desc *desc;
 
@@ -166,7 +166,7 @@ struct irq_desc *irq_desc(unsigned int i
 	}
 	return NULL;
 }
-struct irq_desc *irq_desc_with_new(unsigned int irq)
+struct irq_desc *to_irq_desc_with_new(unsigned int irq)
 {
 	struct irq_desc *desc, *desc_pri;
 	int i;
@@ -245,20 +245,20 @@ struct irq_desc *irq_desc_with_new(unsig
 }
 #else
 
-struct irq_desc *irq_descX;
-DEFINE_DYN_ARRAY(irq_descX, sizeof(struct irq_desc), nr_irqs, PAGE_SIZE, init_work);
+struct irq_desc *irq_desc;
+DEFINE_DYN_ARRAY(irq_desc, sizeof(struct irq_desc), nr_irqs, PAGE_SIZE, init_work);
 
 #endif
 
 #else
 
-struct irq_desc irq_descX[NR_IRQS] __cacheline_aligned_in_smp = {
+struct irq_desc irq_desc[NR_IRQS] __cacheline_aligned_in_smp = {
 	[0 ... NR_IRQS-1] = {
 		.status = IRQ_DISABLED,
 		.chip = &no_irq_chip,
 		.handle_irq = handle_bad_irq,
 		.depth = 1,
-		.lock = __SPIN_LOCK_UNLOCKED(irq_descX->lock),
+		.lock = __SPIN_LOCK_UNLOCKED(irq_desc->lock),
 #ifdef CONFIG_SMP
 		.affinity = CPU_MASK_ALL
 #endif
@@ -268,16 +268,16 @@ struct irq_desc irq_descX[NR_IRQS] __cac
 #endif
 
 #ifndef CONFIG_HAVE_SPARSE_IRQ
-struct irq_desc *irq_desc(unsigned int irq)
+struct irq_desc *to_irq_desc(unsigned int irq)
 {
 	if (irq < nr_irqs)
-		return &irq_descX[irq];
+		return &irq_desc[irq];
 
 	return NULL;
 }
-struct irq_desc *irq_desc_with_new(unsigned int irq)
+struct irq_desc *to_irq_desc_with_new(unsigned int irq)
 {
-	return irq_desc(irq);
+	return to_irq_desc(irq);
 }
 #endif
 
@@ -289,7 +289,7 @@ static void ack_bad(unsigned int irq)
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	print_irq_desc(irq, desc);
 	ack_bad_irq(irq);
 }
@@ -494,7 +494,7 @@ void early_init_irq_lock_class(void)
 
 unsigned int kstat_irqs_cpu(unsigned int irq, int cpu)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	return desc->kstat_irqs[cpu];
 }
 EXPORT_SYMBOL(kstat_irqs_cpu);
Index: linux-2.6/kernel/irq/manage.c
===================================================================
--- linux-2.6.orig/kernel/irq/manage.c
+++ linux-2.6/kernel/irq/manage.c
@@ -31,7 +31,7 @@ cpumask_t irq_default_affinity = CPU_MAS
  */
 void synchronize_irq(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	unsigned int status;
 
 	if (!desc)
@@ -64,7 +64,7 @@ EXPORT_SYMBOL(synchronize_irq);
  */
 int irq_can_set_affinity(unsigned int irq)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	if (CHECK_IRQ_PER_CPU(desc->status) || !desc->chip ||
 	    !desc->chip->set_affinity)
@@ -81,7 +81,7 @@ int irq_can_set_affinity(unsigned int ir
  */
 int irq_set_affinity(unsigned int irq, cpumask_t cpumask)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	if (!desc->chip->set_affinity)
 		return -EINVAL;
@@ -116,7 +116,7 @@ int irq_select_affinity(unsigned int irq
 
 	cpus_and(mask, cpu_online_map, irq_default_affinity);
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	desc->affinity = mask;
 	desc->chip->set_affinity(irq, mask);
 
@@ -142,7 +142,7 @@ void disable_irq_nosync(unsigned int irq
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return;
 
@@ -171,7 +171,7 @@ void disable_irq(unsigned int irq)
 {
 	struct irq_desc *desc;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return;
 
@@ -215,7 +215,7 @@ void enable_irq(unsigned int irq)
 	struct irq_desc *desc;
 	unsigned long flags;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return;
 
@@ -227,7 +227,7 @@ EXPORT_SYMBOL(enable_irq);
 
 int set_irq_wake_real(unsigned int irq, unsigned int on)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	int ret = -ENXIO;
 
 	if (desc->chip->set_wake)
@@ -250,7 +250,7 @@ int set_irq_wake_real(unsigned int irq,
  */
 int set_irq_wake(unsigned int irq, unsigned int on)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	unsigned long flags;
 	int ret = 0;
 
@@ -293,7 +293,7 @@ int can_request_irq(unsigned int irq, un
 	struct irq_desc *desc;
 	struct irqaction *action;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return 0;
 
@@ -357,7 +357,7 @@ int setup_irq(unsigned int irq, struct i
 	int shared = 0;
 	int ret;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return -EINVAL;
 
@@ -513,7 +513,7 @@ void free_irq(unsigned int irq, void *de
 
 	WARN_ON(in_interrupt());
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return;
 
@@ -629,7 +629,7 @@ int request_irq(unsigned int irq, irq_ha
 	if ((irqflags & IRQF_SHARED) && !dev_id)
 		return -EINVAL;
 
-	desc = irq_desc(irq);
+	desc = to_irq_desc(irq);
 	if (!desc)
 		return -EINVAL;
 
Index: linux-2.6/kernel/irq/migration.c
===================================================================
--- linux-2.6.orig/kernel/irq/migration.c
+++ linux-2.6/kernel/irq/migration.c
@@ -3,7 +3,7 @@
 
 void set_pending_irq(unsigned int irq, cpumask_t mask)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	unsigned long flags;
 
 	spin_lock_irqsave(&desc->lock, flags);
Index: linux-2.6/kernel/irq/proc.c
===================================================================
--- linux-2.6.orig/kernel/irq/proc.c
+++ linux-2.6/kernel/irq/proc.c
@@ -19,7 +19,7 @@ static struct proc_dir_entry *root_irq_d
 
 static int irq_affinity_proc_show(struct seq_file *m, void *v)
 {
-	struct irq_desc *desc = irq_desc((long)m->private);
+	struct irq_desc *desc = to_irq_desc((long)m->private);
 	cpumask_t *mask = &desc->affinity;
 
 #ifdef CONFIG_GENERIC_PENDING_IRQ
@@ -43,7 +43,7 @@ static ssize_t irq_affinity_proc_write(s
 	cpumask_t new_value;
 	int err;
 
-	if (!irq_desc(irq)->chip->set_affinity || no_irq_affinity ||
+	if (!to_irq_desc(irq)->chip->set_affinity || no_irq_affinity ||
 	    irq_balancing_disabled(irq))
 		return -EIO;
 
@@ -132,7 +132,7 @@ static const struct file_operations defa
 static int irq_spurious_read(char *page, char **start, off_t off,
 				  int count, int *eof, void *data)
 {
-	struct irq_desc *desc = irq_desc((long) data);
+	struct irq_desc *desc = to_irq_desc((long) data);
 	return sprintf(page, "count %u\n"
 			     "unhandled %u\n"
 			     "last_unhandled %u ms\n",
@@ -145,7 +145,7 @@ static int irq_spurious_read(char *page,
 
 static int name_unique(unsigned int irq, struct irqaction *new_action)
 {
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 	struct irqaction *action;
 	unsigned long flags;
 	int ret = 1;
@@ -165,7 +165,7 @@ static int name_unique(unsigned int irq,
 void register_handler_proc(unsigned int irq, struct irqaction *action)
 {
 	char name [MAX_NAMELEN];
-	struct irq_desc *desc = irq_desc(irq);
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	if (!desc->dir || action->dir || !action->name ||
 					!name_unique(irq, action))
@@ -214,7 +214,7 @@ void register_irq_proc(unsigned int irq,
 void unregister_handler_proc(unsigned int irq, struct irqaction *action)
 {
 	if (action->dir) {
-		struct irq_desc *desc = irq_desc(irq);
+		struct irq_desc *desc = to_irq_desc(irq);
 		remove_proc_entry(action->dir->name, desc->dir);
 	}
 }
Index: linux-2.6/kernel/irq/resend.c
===================================================================
--- linux-2.6.orig/kernel/irq/resend.c
+++ linux-2.6/kernel/irq/resend.c
@@ -36,7 +36,7 @@ static void resend_irqs(unsigned long ar
 	while (!bitmap_empty(irqs_resend, nr_irqs)) {
 		irq = find_first_bit(irqs_resend, nr_irqs);
 		clear_bit(irq, irqs_resend);
-		desc = irq_desc(irq);
+		desc = to_irq_desc(irq);
 		local_irq_disable();
 		desc->handle_irq(irq, desc);
 		local_irq_enable();
Index: linux-2.6/kernel/irq/spurious.c
===================================================================
--- linux-2.6.orig/kernel/irq/spurious.c
+++ linux-2.6/kernel/irq/spurious.c
@@ -97,7 +97,7 @@ static int misrouted_irq(int irq)
 		if (i == irq)	/* Already tried */
 			continue;
 
-		desc = irq_desc(i);
+		desc = to_irq_desc(i);
 		if (try_one_irq(i, desc))
 			ok = 1;
 	}
@@ -109,7 +109,7 @@ static void poll_spurious_irqs(unsigned
 {
 	int i;
 	for (i = 1; i < nr_irqs; i++) {
-		struct irq_desc *desc = irq_desc(i);
+		struct irq_desc *desc = to_irq_desc(i);
 		unsigned int status;
 
 		/* Racy but it doesn't matter */
Index: linux-2.6/arch/mips/sgi-ip22/ip22-int.c
===================================================================
--- linux-2.6.orig/arch/mips/sgi-ip22/ip22-int.c
+++ linux-2.6/arch/mips/sgi-ip22/ip22-int.c
@@ -164,7 +164,7 @@ static void indy_buserror_irq(void)
 	int irq = SGI_BUSERR_IRQ;
 
 	irq_enter();
-	kstat_irqs_this_cpu(irq_desc(irq))++;
+	kstat_irqs_this_cpu(&irq_desc[irq])++;
 	ip22_be_interrupt(irq);
 	irq_exit();
 }
Index: linux-2.6/arch/mips/sgi-ip22/ip22-time.c
===================================================================
--- linux-2.6.orig/arch/mips/sgi-ip22/ip22-time.c
+++ linux-2.6/arch/mips/sgi-ip22/ip22-time.c
@@ -186,7 +186,7 @@ void indy_8254timer_irq(void)
 	char c;
 
 	irq_enter();
-	kstat_irqs_this_cpu(irq_desc(irq))++;
+	kstat_irqs_this_cpu(&irq_desc[irq])++;
 	printk(KERN_ALERT "Oops, got 8254 interrupt.\n");
 	ArcRead(0, &c, 1, &cnt);
 	ArcEnterInteractiveMode();
Index: linux-2.6/arch/mips/sibyte/bcm1480/smp.c
===================================================================
--- linux-2.6.orig/arch/mips/sibyte/bcm1480/smp.c
+++ linux-2.6/arch/mips/sibyte/bcm1480/smp.c
@@ -180,7 +180,7 @@ void bcm1480_mailbox_interrupt(void)
 	int cpu = smp_processor_id();
 	unsigned int action;
 
-	irq_desc(K_BCM1480_INT_MBOX_0_0).kstat_irqs[cpu]++;
+	irq_desc[K_BCM1480_INT_MBOX_0_0].kstat_irqs[cpu]++;
 	/* Load the mailbox register to figure out what we're supposed to do */
 	action = (__raw_readq(mailbox_0_regs[cpu]) >> 48) & 0xffff;
 
Index: linux-2.6/arch/mips/sibyte/sb1250/smp.c
===================================================================
--- linux-2.6.orig/arch/mips/sibyte/sb1250/smp.c
+++ linux-2.6/arch/mips/sibyte/sb1250/smp.c
@@ -168,7 +168,7 @@ void sb1250_mailbox_interrupt(void)
 	int cpu = smp_processor_id();
 	unsigned int action;
 
-	irq_desc(K_INT_MBOX_0)->kstat_irqs[cpu]++;
+	irq_desc[K_INT_MBOX_0].kstat_irqs[cpu]++;
 	/* Load the mailbox register to figure out what we're supposed to do */
 	action = (____raw_readq(mailbox_regs[cpu]) >> 48) & 0xffff;
 
Index: linux-2.6/arch/sparc64/kernel/time.c
===================================================================
--- linux-2.6.orig/arch/sparc64/kernel/time.c
+++ linux-2.6/arch/sparc64/kernel/time.c
@@ -943,7 +943,7 @@ void timer_interrupt(int irq, struct pt_
 
 	irq_enter();
 
-	irq_desc(0)->kstat_irqs[0]++;
+	irq_desc[0].kstat_irqs[0]++;
 
 	if (unlikely(!evt->event_handler)) {
 		printk(KERN_WARNING

Re: [PATCH] x86: apic - unify lapic_resume - fix

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

applied to tip/x86/apic, thanks.

	Ingo

Re: [PATCH 2/2] irq: rename irq_desc() to to_irq_desc()

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> So could revert back to use irq_desc[] for other arch except x86
> 
> Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

great!

could we get rid of the remaining non-x86 arch changes too, please? 

- __do_IRQ(): why is the change needed? Architectures that want to use 
  sparse IRQs should first properly convert to genirq (i.e. should get 
  rid of their __do_IRQ() use).

- kstat_cpu / kstat_irqs_cpu: seems unnecessary now?

- create_irq prototype return value change: cannot we limit the irq 
  space to 0..INT_MAX, and leave the prototype alone?

- create_irq(int) argument change: how about introducing a new 
  create_irq_nr(int) variant, while the create_irq(void) function would 
  just call create_irq_nr(0) ?

	Ingo

diff --git a/arch/alpha/kernel/irq.c b/arch/alpha/kernel/irq.c
index c626a82..b5348ae 100644
--- a/arch/alpha/kernel/irq.c
+++ b/arch/alpha/kernel/irq.c
@@ -89,7 +89,7 @@ show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(irq));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[irq]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(irq, j));
 #endif
 		seq_printf(p, " %14s", irq_desc[irq].chip->typename);
 		seq_printf(p, "  %c%s",
@@ -156,6 +156,6 @@ handle_irq(int irq)
 	 * at IPL 0.
 	 */
 	local_irq_disable();
-	__do_IRQ(irq);
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 }
diff --git a/arch/alpha/kernel/irq_alpha.c b/arch/alpha/kernel/irq_alpha.c
index e16aeb6..2c0852c 100644
--- a/arch/alpha/kernel/irq_alpha.c
+++ b/arch/alpha/kernel/irq_alpha.c
@@ -64,7 +64,7 @@ do_entInt(unsigned long type, unsigned long vector,
 		smp_percpu_timer_interrupt(regs);
 		cpu = smp_processor_id();
 		if (cpu != boot_cpuid) {
-		        kstat_cpu(cpu).irqs[RTC_IRQ]++;
+			irq_desc[RTC_IRQ].kstat_irqs[cpu]++;
 		} else {
 			handle_irq(RTC_IRQ);
 		}
diff --git a/arch/arm/kernel/irq.c b/arch/arm/kernel/irq.c
index f88efb1..26a1185 100644
--- a/arch/arm/kernel/irq.c
+++ b/arch/arm/kernel/irq.c
@@ -69,22 +69,24 @@ int show_interrupts(struct seq_file *p, void *v)
 	}
 
 	if (i < NR_IRQS) {
-		spin_lock_irqsave(&irq_desc[i].lock, flags);
-		action = irq_desc[i].action;
+		struct irq_desc *desc = irq_desc(i);
+
+		spin_lock_irqsave(&desc->lock, flags);
+		action = desc->action;
 		if (!action)
 			goto unlock;
 
 		seq_printf(p, "%3d: ", i);
 		for_each_present_cpu(cpu)
-			seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
-		seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
+		seq_printf(p, " %10s", desc->chip->name ? : "-");
 		seq_printf(p, "  %s", action->name);
 		for (action = action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
 
 		seq_putc(p, '\n');
 unlock:
-		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
+		spin_unlock_irqrestore(&desc->lock, flags);
 	} else if (i == NR_IRQS) {
 #ifdef CONFIG_ARCH_ACORN
 		show_fiq_list(p, v);
diff --git a/arch/arm/mach-iop13xx/msi.c b/arch/arm/mach-iop13xx/msi.c
index 63ef112..ace5bad 100644
--- a/arch/arm/mach-iop13xx/msi.c
+++ b/arch/arm/mach-iop13xx/msi.c
@@ -125,7 +125,7 @@ void __init iop13xx_msi_init(void)
 /*
  * Dynamic irq allocate and deallocation
  */
-int create_irq(void)
+unsigned int create_irq(unsigned not_used)
 {
 	int irq, pos;
 
@@ -173,7 +173,7 @@ static struct irq_chip iop13xx_msi_chip = {
 
 int arch_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *desc)
 {
-	int id, irq = create_irq();
+	int id, irq = create_irq(0);
 	struct msi_msg msg;
 
 	if (irq < 0)
diff --git a/arch/arm/mach-ns9xxx/irq.c b/arch/arm/mach-ns9xxx/irq.c
index 38260d5..5118872 100644
--- a/arch/arm/mach-ns9xxx/irq.c
+++ b/arch/arm/mach-ns9xxx/irq.c
@@ -63,7 +63,6 @@ static struct irq_chip ns9xxx_chip = {
 #else
 static void handle_prio_irq(unsigned int irq, struct irq_desc *desc)
 {
-	unsigned int cpu = smp_processor_id();
 	struct irqaction *action;
 	irqreturn_t action_ret;
 
@@ -72,7 +71,7 @@ static void handle_prio_irq(unsigned int irq, struct irq_desc *desc)
 	BUG_ON(desc->status & IRQ_INPROGRESS);
 
 	desc->status &= ~(IRQ_REPLAY | IRQ_WAITING);
-	kstat_cpu(cpu).irqs[irq]++;
+	kstat_irqs_this_cpu(desc)++;
 
 	action = desc->action;
 	if (unlikely(!action || (desc->status & IRQ_DISABLED)))
diff --git a/arch/avr32/kernel/irq.c b/arch/avr32/kernel/irq.c
index a8e767d..b75a957 100644
--- a/arch/avr32/kernel/irq.c
+++ b/arch/avr32/kernel/irq.c
@@ -51,22 +51,24 @@ int show_interrupts(struct seq_file *p, void *v)
 	}
 
 	if (i < NR_IRQS) {
-		spin_lock_irqsave(&irq_desc[i].lock, flags);
-		action = irq_desc[i].action;
+		struct irq_desc *desc = irq_desc(i);
+
+		spin_lock_irqsave(&desc->lock, flags);
+		action = desc->action;
 		if (!action)
 			goto unlock;
 
 		seq_printf(p, "%3d: ", i);
 		for_each_online_cpu(cpu)
-			seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
-		seq_printf(p, " %8s", irq_desc[i].chip->name ? : "-");
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
+		seq_printf(p, " %8s", desc->chip->name ? : "-");
 		seq_printf(p, "  %s", action->name);
 		for (action = action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
 
 		seq_putc(p, '\n');
 	unlock:
-		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
+		spin_unlock_irqrestore(&desc->lock, flags);
 	}
 
 	return 0;
diff --git a/arch/blackfin/mach-bf527/boards/cm_bf527.c b/arch/blackfin/mach-bf527/boards/cm_bf527.c
index 0b26ae2..08ad8ad 100644
--- a/arch/blackfin/mach-bf527/boards/cm_bf527.c
+++ b/arch/blackfin/mach-bf527/boards/cm_bf527.c
@@ -989,7 +989,7 @@ static int __init stamp_init(void)
 	spi_register_board_info(bfin_spi_board_info, ARRAY_SIZE(bfin_spi_board_info));
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/blackfin/mach-bf527/boards/ezkit.c b/arch/blackfin/mach-bf527/boards/ezkit.c
index 689b69c..66e770d 100644
--- a/arch/blackfin/mach-bf527/boards/ezkit.c
+++ b/arch/blackfin/mach-bf527/boards/ezkit.c
@@ -1024,7 +1024,7 @@ static int __init stamp_init(void)
 	spi_register_board_info(bfin_spi_board_info, ARRAY_SIZE(bfin_spi_board_info));
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/blackfin/mach-bf533/boards/cm_bf533.c b/arch/blackfin/mach-bf533/boards/cm_bf533.c
index ed2b0b8..49b02ec 100644
--- a/arch/blackfin/mach-bf533/boards/cm_bf533.c
+++ b/arch/blackfin/mach-bf533/boards/cm_bf533.c
@@ -418,7 +418,7 @@ static int __init cm_bf533_init(void)
 #endif
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/blackfin/mach-bf533/boards/ezkit.c b/arch/blackfin/mach-bf533/boards/ezkit.c
index 079389c..0b9c3a6 100644
--- a/arch/blackfin/mach-bf533/boards/ezkit.c
+++ b/arch/blackfin/mach-bf533/boards/ezkit.c
@@ -426,7 +426,7 @@ static int __init ezkit_init(void)
 	spi_register_board_info(bfin_spi_board_info, ARRAY_SIZE(bfin_spi_board_info));
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/blackfin/mach-bf533/boards/stamp.c b/arch/blackfin/mach-bf533/boards/stamp.c
index 13ae495..1bb6d18 100644
--- a/arch/blackfin/mach-bf533/boards/stamp.c
+++ b/arch/blackfin/mach-bf533/boards/stamp.c
@@ -627,7 +627,7 @@ static int __init stamp_init(void)
 	spi_register_board_info(bfin_spi_board_info, ARRAY_SIZE(bfin_spi_board_info));
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/blackfin/mach-bf537/boards/generic_board.c b/arch/blackfin/mach-bf537/boards/generic_board.c
index 01b63e2..6f57f09 100644
--- a/arch/blackfin/mach-bf537/boards/generic_board.c
+++ b/arch/blackfin/mach-bf537/boards/generic_board.c
@@ -733,7 +733,7 @@ static int __init stamp_init(void)
 #endif
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/blackfin/mach-bf561/boards/ezkit.c b/arch/blackfin/mach-bf561/boards/ezkit.c
index bc6fede..8cdc0ad 100644
--- a/arch/blackfin/mach-bf561/boards/ezkit.c
+++ b/arch/blackfin/mach-bf561/boards/ezkit.c
@@ -540,7 +540,7 @@ static int __init ezkit_init(void)
 	spi_register_board_info(bfin_spi_board_info, ARRAY_SIZE(bfin_spi_board_info));
 
 #if defined(CONFIG_PATA_PLATFORM) || defined(CONFIG_PATA_PLATFORM_MODULE)
-	irq_desc[PATA_INT].status |= IRQ_NOAUTOEN;
+	irq_desc(PATA_INT)->status |= IRQ_NOAUTOEN;
 #endif
 	return 0;
 }
diff --git a/arch/cris/kernel/irq.c b/arch/cris/kernel/irq.c
index 2dfac8c..a2746de 100644
--- a/arch/cris/kernel/irq.c
+++ b/arch/cris/kernel/irq.c
@@ -66,7 +66,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
@@ -99,7 +99,7 @@ asmlinkage void do_IRQ(int irq, struct pt_regs * regs)
 		printk("do_IRQ: stack overflow: %lX\n", sp);
 		show_stack(NULL, (unsigned long *)sp);
 	}
-	__do_IRQ(irq);
+	__do_IRQ(irq, &irq_desc[irq]);
         irq_exit();
 	set_irq_regs(old_regs);
 }
diff --git a/arch/frv/kernel/irq.c b/arch/frv/kernel/irq.c
index 73abae7..af3e824 100644
--- a/arch/frv/kernel/irq.c
+++ b/arch/frv/kernel/irq.c
@@ -74,7 +74,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		if (action) {
 			seq_printf(p, "%3d: ", i);
 			for_each_present_cpu(cpu)
-				seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
+				seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
 			seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
 			seq_printf(p, "  %s", action->name);
 			for (action = action->next;
diff --git a/arch/h8300/kernel/irq.c b/arch/h8300/kernel/irq.c
index ef4f004..221c4cd 100644
--- a/arch/h8300/kernel/irq.c
+++ b/arch/h8300/kernel/irq.c
@@ -176,7 +176,7 @@ void __init init_IRQ(void)
 asmlinkage void do_IRQ(int irq)
 {
 	irq_enter();
-	__do_IRQ(irq);
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 }
 
@@ -196,7 +196,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		if (!action)
 			goto unlock;
 		seq_printf(p, "%3d: ",i);
-		seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+		seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "-%-8s", irq_desc[i].name);
 		seq_printf(p, "  %s", action->name);
diff --git a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c
index 5c4674a..031e02e 100644
--- a/arch/ia64/kernel/iosapic.c
+++ b/arch/ia64/kernel/iosapic.c
@@ -407,7 +407,7 @@ iosapic_end_level_irq (unsigned int irq)
 		iosapic_eoi(rte->iosapic->addr, vec);
 
 	if (unlikely(do_unmask_irq)) {
-		move_masked_irq(irq);
+		move_masked_irq(irq, &irq_desc[irq]);
 		unmask_irq(irq);
 	}
 }
@@ -452,7 +452,7 @@ iosapic_ack_edge_irq (unsigned int irq)
 	irq_desc_t *idesc = irq_desc + irq;
 
 	irq_complete_move(irq);
-	move_native_irq(irq);
+	move_native_irq(irq, idesc);
 	/*
 	 * Once we have recorded IRQ_PENDING already, we can mask the
 	 * interrupt for real. This prevents IRQ storms from unhandled
@@ -532,7 +532,7 @@ iosapic_reassign_vector (int irq)
 	int new_irq;
 
 	if (iosapic_intr_info[irq].count) {
-		new_irq = create_irq();
+		new_irq = create_irq(0);
 		if (new_irq < 0)
 			panic("%s: out of interrupt vectors!\n", __func__);
 		printk(KERN_INFO "Reassigning vector %d to %d\n",
@@ -783,7 +783,7 @@ iosapic_register_intr (unsigned int gsi,
 			goto unlock_iosapic_lock;
 		}
 	} else
-		irq = create_irq();
+		irq = create_irq(0);
 
 	/* If vector is running out, we try to find a sharable vector */
 	if (irq < 0) {
@@ -922,7 +922,7 @@ iosapic_register_platform_intr (u32 int_type, unsigned int gsi,
 		delivery = IOSAPIC_PMI;
 		break;
 	      case ACPI_INTERRUPT_INIT:
-		irq = create_irq();
+		irq = create_irq(0);
 		if (irq < 0)
 			panic("%s: out of interrupt vectors!\n", __func__);
 		vector = irq_to_vector(irq);
diff --git a/arch/ia64/kernel/irq.c b/arch/ia64/kernel/irq.c
index 7fd18f5..4140d51 100644
--- a/arch/ia64/kernel/irq.c
+++ b/arch/ia64/kernel/irq.c
@@ -80,7 +80,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j) {
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 		}
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
diff --git a/arch/ia64/kernel/irq_ia64.c b/arch/ia64/kernel/irq_ia64.c
index 28d3d48..96f7571 100644
--- a/arch/ia64/kernel/irq_ia64.c
+++ b/arch/ia64/kernel/irq_ia64.c
@@ -403,7 +403,7 @@ void destroy_and_reserve_irq(unsigned int irq)
 /*
  * Dynamic irq allocate and deallocation for MSI
  */
-int create_irq(void)
+unsigned int create_irq(unsigned int not_used)
 {
 	unsigned long flags;
 	int irq, vector, cpu;
@@ -495,9 +495,9 @@ ia64_handle_irq (ia64_vector vector, struct pt_regs *regs)
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
 		if (unlikely(IS_LOCAL_TLB_FLUSH(vector))) {
 			smp_local_flush_tlb();
-			kstat_this_cpu.irqs[vector]++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		} else if (unlikely(IS_RESCHEDULE(vector)))
-			kstat_this_cpu.irqs[vector]++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		else {
 			int irq = local_vector_to_irq(vector);
 
@@ -553,9 +553,9 @@ void ia64_process_pending_intr(void)
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
 		if (unlikely(IS_LOCAL_TLB_FLUSH(vector))) {
 			smp_local_flush_tlb();
-			kstat_this_cpu.irqs[vector]++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		} else if (unlikely(IS_RESCHEDULE(vector)))
-			kstat_this_cpu.irqs[vector]++;
+			kstat_irqs_this_cpu(&irq_desc[vector])++;
 		else {
 			struct pt_regs *old_regs = set_irq_regs(NULL);
 			int irq = local_vector_to_irq(vector);
diff --git a/arch/ia64/kernel/msi_ia64.c b/arch/ia64/kernel/msi_ia64.c
index 60c6ef6..924293f 100644
--- a/arch/ia64/kernel/msi_ia64.c
+++ b/arch/ia64/kernel/msi_ia64.c
@@ -84,7 +84,7 @@ int ia64_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *desc)
 	int	irq, vector;
 	cpumask_t mask;
 
-	irq = create_irq();
+	irq = create_irq(0);
 	if (irq < 0)
 		return irq;
 
@@ -120,7 +120,7 @@ void ia64_teardown_msi_irq(unsigned int irq)
 static void ia64_ack_msi_irq(unsigned int irq)
 {
 	irq_complete_move(irq);
-	move_native_irq(irq);
+	move_native_irq(irq, &irq_desc[irq]);
 	ia64_eoi();
 }
 
diff --git a/arch/ia64/sn/kernel/irq.c b/arch/ia64/sn/kernel/irq.c
index 96c31b4..50ddcf3 100644
--- a/arch/ia64/sn/kernel/irq.c
+++ b/arch/ia64/sn/kernel/irq.c
@@ -110,7 +110,7 @@ static void sn_ack_irq(unsigned int irq)
 	HUB_S((u64*)LOCAL_MMR_ADDR(SH_EVENT_OCCURRED_ALIAS), mask);
 	__set_bit(irq, (volatile void *)pda->sn_in_service_ivecs);
 
-	move_native_irq(irq);
+	move_native_irq(irq, &irq_desc[irq]);
 }
 
 static void sn_end_irq(unsigned int irq)
diff --git a/arch/ia64/sn/kernel/msi_sn.c b/arch/ia64/sn/kernel/msi_sn.c
index 83f190f..a6739fa 100644
--- a/arch/ia64/sn/kernel/msi_sn.c
+++ b/arch/ia64/sn/kernel/msi_sn.c
@@ -83,7 +83,7 @@ int sn_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *entry)
 	if (provider == NULL || provider->dma_map_consistent == NULL)
 		return -EINVAL;
 
-	irq = create_irq();
+	irq = create_irq(0);
 	if (irq < 0)
 		return irq;
 
@@ -210,7 +210,7 @@ static void sn_set_msi_irq_affinity(unsigned int irq, cpumask_t cpu_mask)
 
 static void sn_ack_msi_irq(unsigned int irq)
 {
-	move_native_irq(irq);
+	move_native_irq(irq, &irq_desc[irq]);
 	ia64_eoi();
 }
 
diff --git a/arch/m32r/kernel/irq.c b/arch/m32r/kernel/irq.c
index d0c5b0b..09d16ea 100644
--- a/arch/m32r/kernel/irq.c
+++ b/arch/m32r/kernel/irq.c
@@ -52,7 +52,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
@@ -84,7 +84,7 @@ asmlinkage unsigned int do_IRQ(int irq, struct pt_regs *regs)
 #ifdef CONFIG_DEBUG_STACKOVERFLOW
 	/* FIXME M32R */
 #endif
-	__do_IRQ(irq);
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 	set_irq_regs(old_regs);
 
diff --git a/arch/m68k/kernel/ints.c b/arch/m68k/kernel/ints.c
index ded7dd2..d8aebfc 100644
--- a/arch/m68k/kernel/ints.c
+++ b/arch/m68k/kernel/ints.c
@@ -46,6 +46,8 @@
 #include <asm/q40ints.h>
 #endif
 
+int nr_irqs = NR_IRQS;
+
 extern u32 auto_irqhandler_fixup[];
 extern u32 user_irqhandler_fixup[];
 extern u16 user_irqvec_fixup[];
diff --git a/arch/m68knommu/kernel/irq.c b/arch/m68knommu/kernel/irq.c
index bba1bb4..7d565f1 100644
--- a/arch/m68knommu/kernel/irq.c
+++ b/arch/m68knommu/kernel/irq.c
@@ -23,7 +23,7 @@ asmlinkage void do_IRQ(int irq, struct pt_regs *regs)
 	struct pt_regs *oldregs = set_irq_regs(regs);
 
 	irq_enter();
-	__do_IRQ(irq);
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 
 	set_irq_regs(oldregs);
diff --git a/arch/mips/kernel/irq.c b/arch/mips/kernel/irq.c
index 4b4007b..7b845ba 100644
--- a/arch/mips/kernel/irq.c
+++ b/arch/mips/kernel/irq.c
@@ -108,7 +108,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "  %s", action->name);
diff --git a/arch/mips/sgi-ip22/ip22-int.c b/arch/mips/sgi-ip22/ip22-int.c
index f6d9bf4..6794c76 100644
--- a/arch/mips/sgi-ip22/ip22-int.c
+++ b/arch/mips/sgi-ip22/ip22-int.c
@@ -164,7 +164,7 @@ static void indy_buserror_irq(void)
 	int irq = SGI_BUSERR_IRQ;
 
 	irq_enter();
-	kstat_this_cpu.irqs[irq]++;
+	kstat_irqs_this_cpu(&irq_desc[irq])++;
 	ip22_be_interrupt(irq);
 	irq_exit();
 }
diff --git a/arch/mips/sgi-ip22/ip22-time.c b/arch/mips/sgi-ip22/ip22-time.c
index 10e5054..f744b2e 100644
--- a/arch/mips/sgi-ip22/ip22-time.c
+++ b/arch/mips/sgi-ip22/ip22-time.c
@@ -186,7 +186,7 @@ void indy_8254timer_irq(void)
 	char c;
 
 	irq_enter();
-	kstat_this_cpu.irqs[irq]++;
+	kstat_irqs_this_cpu(&irq_desc[irq])++;
 	printk(KERN_ALERT "Oops, got 8254 interrupt.\n");
 	ArcRead(0, &c, 1, &cnt);
 	ArcEnterInteractiveMode();
diff --git a/arch/mips/sibyte/bcm1480/smp.c b/arch/mips/sibyte/bcm1480/smp.c
index bd9eeb4..6f09896 100644
--- a/arch/mips/sibyte/bcm1480/smp.c
+++ b/arch/mips/sibyte/bcm1480/smp.c
@@ -180,7 +180,7 @@ void bcm1480_mailbox_interrupt(void)
 	int cpu = smp_processor_id();
 	unsigned int action;
 
-	kstat_this_cpu.irqs[K_BCM1480_INT_MBOX_0_0]++;
+	irq_desc[K_BCM1480_INT_MBOX_0_0].kstat_irqs[cpu]++;
 	/* Load the mailbox register to figure out what we're supposed to do */
 	action = (__raw_readq(mailbox_0_regs[cpu]) >> 48) & 0xffff;
 
diff --git a/arch/mips/sibyte/sb1250/smp.c b/arch/mips/sibyte/sb1250/smp.c
index 0734b93..dd0e55e 100644
--- a/arch/mips/sibyte/sb1250/smp.c
+++ b/arch/mips/sibyte/sb1250/smp.c
@@ -168,7 +168,7 @@ void sb1250_mailbox_interrupt(void)
 	int cpu = smp_processor_id();
 	unsigned int action;
 
-	kstat_this_cpu.irqs[K_INT_MBOX_0]++;
+	irq_desc[K_INT_MBOX_0].kstat_irqs[cpu]++;
 	/* Load the mailbox register to figure out what we're supposed to do */
 	action = (____raw_readq(mailbox_regs[cpu]) >> 48) & 0xffff;
 
diff --git a/arch/mn10300/kernel/irq.c b/arch/mn10300/kernel/irq.c
index 761c434..87ecf9d 100644
--- a/arch/mn10300/kernel/irq.c
+++ b/arch/mn10300/kernel/irq.c
@@ -197,12 +197,11 @@ int show_interrupts(struct seq_file *p, void *v)
 		/* display information rows, one per active CPU */
 	case 1 ... NR_IRQS - 1:
 		spin_lock_irqsave(&irq_desc[i].lock, flags);
-
 		action = irq_desc[i].action;
 		if (action) {
 			seq_printf(p, "%3d: ", i);
 			for_each_present_cpu(cpu)
-				seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
+				seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
 			seq_printf(p, " %14s.%u", irq_desc[i].chip->name,
 				   (GxICR(i) & GxICR_LEVEL) >>
 				   GxICR_LEVEL_SHIFT);
diff --git a/arch/parisc/kernel/irq.c b/arch/parisc/kernel/irq.c
index 23ef950..8672b27 100644
--- a/arch/parisc/kernel/irq.c
+++ b/arch/parisc/kernel/irq.c
@@ -183,7 +183,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%3d: ", i);
 #ifdef CONFIG_SMP
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #else
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #endif
@@ -363,7 +363,7 @@ void do_cpu_irq_mask(struct pt_regs *regs)
 		goto set_out;
 	}
 #endif
-	__do_IRQ(irq);
+	__do_IRQ(irq, irq_desc);
 
  out:
 	irq_exit();
diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c
index d972dec..93870a6 100644
--- a/arch/powerpc/kernel/irq.c
+++ b/arch/powerpc/kernel/irq.c
@@ -190,7 +190,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%3d: ", i);
 #ifdef CONFIG_SMP
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #else
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #endif /* CONFIG_SMP */
diff --git a/arch/powerpc/platforms/cell/interrupt.c b/arch/powerpc/platforms/cell/interrupt.c
index 2d5bb22..490b619 100644
--- a/arch/powerpc/platforms/cell/interrupt.c
+++ b/arch/powerpc/platforms/cell/interrupt.c
@@ -254,7 +254,7 @@ static void handle_iic_irq(unsigned int irq, struct irq_desc *desc)
 		goto out_eoi;
 	}
 
-	kstat_cpu(cpu).irqs[irq]++;
+	kstat_irqs_this_cpu(desc)++;
 
 	/* Mark the IRQ currently in progress.*/
 	desc->status |= IRQ_INPROGRESS;
diff --git a/arch/powerpc/platforms/powermac/pic.c b/arch/powerpc/platforms/powermac/pic.c
index 6d149ae..e3d44d5 100644
--- a/arch/powerpc/platforms/powermac/pic.c
+++ b/arch/powerpc/platforms/powermac/pic.c
@@ -221,7 +221,7 @@ static irqreturn_t gatwick_action(int cpl, void *dev_id)
 			continue;
 		irq += __ilog2(bits);
 		spin_unlock_irqrestore(&pmac_pic_lock, flags);
-		__do_IRQ(irq);
+		__do_IRQ(irq, &irq_desc[irq]);
 		spin_lock_irqsave(&pmac_pic_lock, flags);
 		rc = IRQ_HANDLED;
 	}
diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c
index e7c5bfb..14eb549 100644
--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -17,6 +17,8 @@
 #include <linux/proc_fs.h>
 #include <linux/profile.h>
 
+int nr_irqs = NR_IRQS;
+
 /*
  * show_interrupts is needed by /proc/interrupts.
  */
diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c
index 64b7690..0080a16 100644
--- a/arch/sh/kernel/irq.c
+++ b/arch/sh/kernel/irq.c
@@ -51,7 +51,7 @@ int show_interrupts(struct seq_file *p, void *v)
 			goto unlock;
 		seq_printf(p, "%3d: ",i);
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "-%-8s", irq_desc[i].name);
 		seq_printf(p, "  %s", action->name);
diff --git a/arch/sparc/kernel/irq.c b/arch/sparc/kernel/irq.c
index 93e1d1c..059598b 100644
--- a/arch/sparc/kernel/irq.c
+++ b/arch/sparc/kernel/irq.c
@@ -55,6 +55,9 @@
 #define SMP_NOP2
 #define SMP_NOP3
 #endif /* SMP */
+
+int nr_irqs = NR_IRQS;
+
 unsigned long __raw_local_irq_save(void)
 {
 	unsigned long retval;
diff --git a/arch/sparc64/kernel/irq.c b/arch/sparc64/kernel/irq.c
index 9b6689d..fc0fd8b 100644
--- a/arch/sparc64/kernel/irq.c
+++ b/arch/sparc64/kernel/irq.c
@@ -185,7 +185,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
 		seq_printf(p, " %9s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
diff --git a/arch/sparc64/kernel/time.c b/arch/sparc64/kernel/time.c
index cc16fdc..db211ac 100644
--- a/arch/sparc64/kernel/time.c
+++ b/arch/sparc64/kernel/time.c
@@ -16,6 +16,7 @@
 #include <linux/param.h>
 #include <linux/string.h>
 #include <linux/mm.h>
+#include <linux/irq.h>
 #include <linux/interrupt.h>
 #include <linux/time.h>
 #include <linux/timex.h>
@@ -942,7 +943,7 @@ void timer_interrupt(int irq, struct pt_regs *regs)
 
 	irq_enter();
 
-	kstat_this_cpu.irqs[0]++;
+	irq_desc[0].kstat_irqs[0]++;
 
 	if (unlikely(!evt->event_handler)) {
 		printk(KERN_WARNING
diff --git a/arch/um/kernel/irq.c b/arch/um/kernel/irq.c
index 3d7aad0..a2ff0ea 100644
--- a/arch/um/kernel/irq.c
+++ b/arch/um/kernel/irq.c
@@ -42,7 +42,7 @@ int show_interrupts(struct seq_file *p, void *v)
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
@@ -332,7 +332,7 @@ unsigned int do_IRQ(int irq, struct uml_pt_regs *regs)
 {
 	struct pt_regs *old_regs = set_irq_regs((struct pt_regs *)regs);
 	irq_enter();
-	__do_IRQ(irq);
+	__do_IRQ(irq, &irq_desc[irq]);
 	irq_exit();
 	set_irq_regs(old_regs);
 	return 1;

Re: [PATCH 2/2] irq: rename irq_desc() to to_irq_desc()

[Yinghai Lu]

[-- Attachment #1: Type: text/plain, Size: 947 bytes --]

On Mon, Aug 18, 2008 at 12:37 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>
>> So could revert back to use irq_desc[] for other arch except x86
>>
>> Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
>
> great!
>
> could we get rid of the remaining non-x86 arch changes too, please?
>
> - __do_IRQ(): why is the change needed? Architectures that want to use
>  sparse IRQs should first properly convert to genirq (i.e. should get
>  rid of their __do_IRQ() use).

please check attacch patch

>
> - kstat_cpu / kstat_irqs_cpu: seems unnecessary now?

that is needed

>
> - create_irq prototype return value change: cannot we limit the irq
>  space to 0..INT_MAX, and leave the prototype alone?
>
> - create_irq(int) argument change: how about introducing a new
>  create_irq_nr(int) variant, while the create_irq(void) function would
>  just call create_irq_nr(0) ?

will send out another patch later

YH

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: irq_desc__to_irq_desc_fix1.patch --]
[-- Type: text/x-patch; name=irq_desc__to_irq_desc_fix1.patch, Size: 13889 bytes --]

[PATCH] irq: rename irq_desc() to to_irq_desc() - fix

some left overs.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/alpha/kernel/irq.c               |    2 +-
 arch/arm/kernel/irq.c                 |   10 ++++------
 arch/avr32/kernel/irq.c               |    8 +++-----
 arch/cris/kernel/irq.c                |    2 +-
 arch/h8300/kernel/irq.c               |    2 +-
 arch/ia64/kernel/iosapic.c            |    2 +-
 arch/ia64/kernel/msi_ia64.c           |    2 +-
 arch/ia64/sn/kernel/irq.c             |    2 +-
 arch/ia64/sn/kernel/msi_sn.c          |    2 +-
 arch/m32r/kernel/irq.c                |    2 +-
 arch/m68knommu/kernel/irq.c           |    2 +-
 arch/parisc/kernel/irq.c              |    2 +-
 arch/powerpc/platforms/powermac/pic.c |    2 +-
 arch/um/kernel/irq.c                  |    2 +-
 arch/x86/kernel/io_apic.c             |    4 +---
 drivers/parisc/dino.c                 |    3 +--
 drivers/parisc/eisa.c                 |    4 +---
 drivers/parisc/gsc.c                  |    3 +--
 drivers/parisc/superio.c              |    4 +---
 drivers/xen/events.c                  |    2 +-
 include/linux/irq.h                   |   10 +++++-----
 kernel/irq/handle.c                   |    3 ++-
 kernel/irq/migration.c                |    4 +++-
 23 files changed, 35 insertions(+), 44 deletions(-)

Index: linux-2.6/arch/alpha/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq.c
+++ linux-2.6/arch/alpha/kernel/irq.c
@@ -156,6 +156,6 @@ handle_irq(int irq)
 	 * at IPL 0.
 	 */
 	local_irq_disable();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 }
Index: linux-2.6/arch/arm/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/kernel/irq.c
+++ linux-2.6/arch/arm/kernel/irq.c
@@ -69,24 +69,22 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 
 		seq_printf(p, "%3d: ", i);
 		for_each_present_cpu(cpu)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
-		seq_printf(p, " %10s", desc->chip->name ? : "-");
+		seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
 		seq_printf(p, "  %s", action->name);
 		for (action = action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
 
 		seq_putc(p, '\n');
 unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 #ifdef CONFIG_ARCH_ACORN
 		show_fiq_list(p, v);
Index: linux-2.6/arch/avr32/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/avr32/kernel/irq.c
+++ linux-2.6/arch/avr32/kernel/irq.c
@@ -51,10 +51,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 
@@ -68,7 +66,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 	unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	}
 
 	return 0;
Index: linux-2.6/arch/cris/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/cris/kernel/irq.c
+++ linux-2.6/arch/cris/kernel/irq.c
@@ -99,7 +99,7 @@ asmlinkage void do_IRQ(int irq, struct p
 		printk("do_IRQ: stack overflow: %lX\n", sp);
 		show_stack(NULL, (unsigned long *)sp);
 	}
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
         irq_exit();
 	set_irq_regs(old_regs);
 }
Index: linux-2.6/arch/h8300/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/h8300/kernel/irq.c
+++ linux-2.6/arch/h8300/kernel/irq.c
@@ -176,7 +176,7 @@ void __init init_IRQ(void)
 asmlinkage void do_IRQ(int irq)
 {
 	irq_enter();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 }
 
Index: linux-2.6/arch/ia64/kernel/iosapic.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/iosapic.c
+++ linux-2.6/arch/ia64/kernel/iosapic.c
@@ -452,7 +452,7 @@ iosapic_ack_edge_irq (unsigned int irq)
 	irq_desc_t *idesc = irq_desc + irq;
 
 	irq_complete_move(irq);
-	move_native_irq(irq, idesc);
+	move_native_irq(irq);
 	/*
 	 * Once we have recorded IRQ_PENDING already, we can mask the
 	 * interrupt for real. This prevents IRQ storms from unhandled
Index: linux-2.6/arch/ia64/kernel/msi_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/msi_ia64.c
+++ linux-2.6/arch/ia64/kernel/msi_ia64.c
@@ -120,7 +120,7 @@ void ia64_teardown_msi_irq(unsigned int
 static void ia64_ack_msi_irq(unsigned int irq)
 {
 	irq_complete_move(irq);
-	move_native_irq(irq, &irq_desc[irq]);
+	move_native_irq(irq);
 	ia64_eoi();
 }
 
Index: linux-2.6/arch/ia64/sn/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/irq.c
+++ linux-2.6/arch/ia64/sn/kernel/irq.c
@@ -110,7 +110,7 @@ static void sn_ack_irq(unsigned int irq)
 	HUB_S((u64*)LOCAL_MMR_ADDR(SH_EVENT_OCCURRED_ALIAS), mask);
 	__set_bit(irq, (volatile void *)pda->sn_in_service_ivecs);
 
-	move_native_irq(irq, &irq_desc[irq]);
+	move_native_irq(irq);
 }
 
 static void sn_end_irq(unsigned int irq)
Index: linux-2.6/arch/ia64/sn/kernel/msi_sn.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/msi_sn.c
+++ linux-2.6/arch/ia64/sn/kernel/msi_sn.c
@@ -210,7 +210,7 @@ static void sn_set_msi_irq_affinity(unsi
 
 static void sn_ack_msi_irq(unsigned int irq)
 {
-	move_native_irq(irq, &irq_desc[irq]);
+	move_native_irq(irq);
 	ia64_eoi();
 }
 
Index: linux-2.6/arch/m32r/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m32r/kernel/irq.c
+++ linux-2.6/arch/m32r/kernel/irq.c
@@ -84,7 +84,7 @@ asmlinkage unsigned int do_IRQ(int irq,
 #ifdef CONFIG_DEBUG_STACKOVERFLOW
 	/* FIXME M32R */
 #endif
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 	set_irq_regs(old_regs);
 
Index: linux-2.6/arch/m68knommu/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m68knommu/kernel/irq.c
+++ linux-2.6/arch/m68knommu/kernel/irq.c
@@ -23,7 +23,7 @@ asmlinkage void do_IRQ(int irq, struct p
 	struct pt_regs *oldregs = set_irq_regs(regs);
 
 	irq_enter();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 
 	set_irq_regs(oldregs);
Index: linux-2.6/arch/parisc/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/parisc/kernel/irq.c
+++ linux-2.6/arch/parisc/kernel/irq.c
@@ -363,7 +363,7 @@ void do_cpu_irq_mask(struct pt_regs *reg
 		goto set_out;
 	}
 #endif
-	__do_IRQ(irq, irq_desc);
+	__do_IRQ(irq);
 
  out:
 	irq_exit();
Index: linux-2.6/arch/powerpc/platforms/powermac/pic.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/powermac/pic.c
+++ linux-2.6/arch/powerpc/platforms/powermac/pic.c
@@ -221,7 +221,7 @@ static irqreturn_t gatwick_action(int cp
 			continue;
 		irq += __ilog2(bits);
 		spin_unlock_irqrestore(&pmac_pic_lock, flags);
-		__do_IRQ(irq, &irq_desc[irq]);
+		__do_IRQ(irq);
 		spin_lock_irqsave(&pmac_pic_lock, flags);
 		rc = IRQ_HANDLED;
 	}
Index: linux-2.6/arch/um/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/um/kernel/irq.c
+++ linux-2.6/arch/um/kernel/irq.c
@@ -332,7 +332,7 @@ unsigned int do_IRQ(int irq, struct uml_
 {
 	struct pt_regs *old_regs = set_irq_regs((struct pt_regs *)regs);
 	irq_enter();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 	set_irq_regs(old_regs);
 	return 1;
Index: linux-2.6/drivers/parisc/dino.c
===================================================================
--- linux-2.6.orig/drivers/parisc/dino.c
+++ linux-2.6/drivers/parisc/dino.c
@@ -389,10 +389,9 @@ ilr_again:
 	do {
 		int local_irq = __ffs(mask);
 		int irq = dino_dev->global_irq[local_irq];
-		struct irq_desc *desc = to_irq_desc(irq);
 		DBG(KERN_DEBUG "%s(%d, %p) mask 0x%x\n",
 			__func__, irq, intr_dev, mask);
-		__do_IRQ(irq, desc);
+		__do_IRQ(irq);
 		mask &= ~(1 << local_irq);
 	} while (mask);
 
Index: linux-2.6/drivers/parisc/eisa.c
===================================================================
--- linux-2.6.orig/drivers/parisc/eisa.c
+++ linux-2.6/drivers/parisc/eisa.c
@@ -202,7 +202,6 @@ static irqreturn_t eisa_irq(int wax_irq,
 {
 	int irq = gsc_readb(0xfc01f000); /* EISA supports 16 irqs */
 	unsigned long flags;
-	struct irq_desc *desc;
         
 	spin_lock_irqsave(&eisa_irq_lock, flags);
 	/* read IRR command */
@@ -234,8 +233,7 @@ static irqreturn_t eisa_irq(int wax_irq,
 	}
 	spin_unlock_irqrestore(&eisa_irq_lock, flags);
 
-	desc = to_irq_desc(irq);
-	__do_IRQ(irq, desc);
+	__do_IRQ(irq);
    
 	spin_lock_irqsave(&eisa_irq_lock, flags);
 	/* unmask */
Index: linux-2.6/drivers/parisc/gsc.c
===================================================================
--- linux-2.6.orig/drivers/parisc/gsc.c
+++ linux-2.6/drivers/parisc/gsc.c
@@ -87,8 +87,7 @@ irqreturn_t gsc_asic_intr(int gsc_asic_i
 	do {
 		int local_irq = __ffs(irr);
 		unsigned int irq = gsc_asic->global_irq[local_irq];
-		struct irq_desc *desc = to_irq_desc(irq);
-		__do_IRQ(irq, desc);
+		__do_IRQ(irq);
 		irr &= ~(1 << local_irq);
 	} while (irr);
 
Index: linux-2.6/drivers/parisc/superio.c
===================================================================
--- linux-2.6.orig/drivers/parisc/superio.c
+++ linux-2.6/drivers/parisc/superio.c
@@ -99,7 +99,6 @@ superio_interrupt(int parent_irq, void *
 {
 	u8 results;
 	u8 local_irq;
-	struct irq_desc *desc;
 
 	/* Poll the 8259 to see if there's an interrupt. */
 	outb (OCW3_POLL,IC_PIC1+0);
@@ -140,8 +139,7 @@ superio_interrupt(int parent_irq, void *
 	}
 
 	/* Call the appropriate device's interrupt */
-	desc = to_irq_desc(local_irq);
-	__do_IRQ(local_irq, desc);
+	__do_IRQ(local_irq);
 
 	/* set EOI - forces a new interrupt if a lower priority device
 	 * still needs service.
Index: linux-2.6/drivers/xen/events.c
===================================================================
--- linux-2.6.orig/drivers/xen/events.c
+++ linux-2.6/drivers/xen/events.c
@@ -640,7 +640,7 @@ static void ack_dynirq(unsigned int irq)
 {
 	int evtchn = evtchn_from_irq(irq);
 
-	move_native_irq(irq, to_irq_desc(irq));
+	move_native_irq(irq);
 
 	if (VALID_EVTCHN(evtchn))
 		clear_evtchn(evtchn);
Index: linux-2.6/include/linux/irq.h
===================================================================
--- linux-2.6.orig/include/linux/irq.h
+++ linux-2.6/include/linux/irq.h
@@ -244,7 +244,7 @@ extern int setup_irq(unsigned int irq, s
 #ifdef CONFIG_GENERIC_PENDING_IRQ
 
 void set_pending_irq(unsigned int irq, cpumask_t mask);
-void move_native_irq(int irq, struct irq_desc *desc);
+void move_native_irq(int irq);
 void move_masked_irq(int irq, struct irq_desc *desc);
 
 #else /* CONFIG_GENERIC_PENDING_IRQ */
@@ -253,7 +253,7 @@ static inline void move_irq(int irq)
 {
 }
 
-static inline void move_native_irq(int irq, struct irq_desc *desc)
+static inline void move_native_irq(int irq)
 {
 }
 
@@ -269,7 +269,7 @@ static inline void set_pending_irq(unsig
 
 #else /* CONFIG_SMP */
 
-#define move_native_irq(x, y)
+#define move_native_irq(x)
 #define move_masked_irq(x, y)
 
 #endif /* CONFIG_SMP */
@@ -302,7 +302,7 @@ extern void handle_bad_irq(unsigned int
  * Monolithic do_IRQ implementation.
  */
 #ifndef CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ
-extern unsigned int __do_IRQ(unsigned int irq, struct irq_desc *desc);
+extern unsigned int __do_IRQ(unsigned int irq);
 #endif
 
 /*
@@ -319,7 +319,7 @@ static inline void generic_handle_irq_de
 	if (likely(desc->handle_irq))
 		desc->handle_irq(irq, desc);
 	else
-		__do_IRQ(irq, desc);
+		__do_IRQ(irq);
 #endif
 }
 
Index: linux-2.6/kernel/irq/handle.c
===================================================================
--- linux-2.6.orig/kernel/irq/handle.c
+++ linux-2.6/kernel/irq/handle.c
@@ -385,10 +385,11 @@ irqreturn_t handle_IRQ_event(unsigned in
  * This is the original x86 implementation which is used for every
  * interrupt type.
  */
-unsigned int __do_IRQ(unsigned int irq, struct irq_desc *desc)
+unsigned int __do_IRQ(unsigned int irq)
 {
 	struct irqaction *action;
 	unsigned int status;
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	kstat_irqs_this_cpu(desc)++;
 	if (CHECK_IRQ_PER_CPU(desc->status)) {
Index: linux-2.6/kernel/irq/migration.c
===================================================================
--- linux-2.6.orig/kernel/irq/migration.c
+++ linux-2.6/kernel/irq/migration.c
@@ -57,8 +57,10 @@ void move_masked_irq(int irq, struct irq
 	cpus_clear(desc->pending_mask);
 }
 
-void move_native_irq(int irq, struct irq_desc *desc)
+void move_native_irq(int irq)
 {
+	struct irq_desc *desc = to_irq_desc(irq);
+
 	if (likely(!(desc->status & IRQ_MOVE_PENDING)))
 		return;
 
Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic.c
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -2426,10 +2426,8 @@ static void ack_x2apic_edge(unsigned int
 
 static void ack_apic_edge(unsigned int irq)
 {
-	struct irq_desc *desc = to_irq_desc(irq);
-
 	irq_complete_move(irq);
-	move_native_irq(irq, desc);
+	move_native_irq(irq);
 	ack_APIC_irq();
 }
 

[PATCH] irq: rename irq_desc() to to_irq_desc() - fix

[Yinghai Lu]

some left overs.
__do_IRQ not take desc as parameter

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/alpha/kernel/irq.c               |    2 +-
 arch/arm/kernel/irq.c                 |   10 ++++------
 arch/avr32/kernel/irq.c               |    8 +++-----
 arch/cris/kernel/irq.c                |    2 +-
 arch/h8300/kernel/irq.c               |    2 +-
 arch/ia64/kernel/iosapic.c            |    2 +-
 arch/ia64/kernel/msi_ia64.c           |    2 +-
 arch/ia64/sn/kernel/irq.c             |    2 +-
 arch/ia64/sn/kernel/msi_sn.c          |    2 +-
 arch/m32r/kernel/irq.c                |    2 +-
 arch/m68knommu/kernel/irq.c           |    2 +-
 arch/parisc/kernel/irq.c              |    2 +-
 arch/powerpc/platforms/powermac/pic.c |    2 +-
 arch/um/kernel/irq.c                  |    2 +-
 arch/x86/kernel/io_apic.c             |    4 +---
 drivers/parisc/dino.c                 |    3 +--
 drivers/parisc/eisa.c                 |    4 +---
 drivers/parisc/gsc.c                  |    3 +--
 drivers/parisc/superio.c              |    4 +---
 drivers/xen/events.c                  |    2 +-
 include/linux/irq.h                   |   10 +++++-----
 kernel/irq/handle.c                   |    3 ++-
 kernel/irq/migration.c                |    4 +++-
 23 files changed, 35 insertions(+), 44 deletions(-)

Index: linux-2.6/arch/alpha/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq.c
+++ linux-2.6/arch/alpha/kernel/irq.c
@@ -156,6 +156,6 @@ handle_irq(int irq)
 	 * at IPL 0.
 	 */
 	local_irq_disable();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 }
Index: linux-2.6/arch/arm/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/kernel/irq.c
+++ linux-2.6/arch/arm/kernel/irq.c
@@ -69,24 +69,22 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 
 		seq_printf(p, "%3d: ", i);
 		for_each_present_cpu(cpu)
 			seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
-		seq_printf(p, " %10s", desc->chip->name ? : "-");
+		seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
 		seq_printf(p, "  %s", action->name);
 		for (action = action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
 
 		seq_putc(p, '\n');
 unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	} else if (i == NR_IRQS) {
 #ifdef CONFIG_ARCH_ACORN
 		show_fiq_list(p, v);
Index: linux-2.6/arch/avr32/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/avr32/kernel/irq.c
+++ linux-2.6/arch/avr32/kernel/irq.c
@@ -51,10 +51,8 @@ int show_interrupts(struct seq_file *p,
 	}
 
 	if (i < NR_IRQS) {
-		struct irq_desc *desc = irq_desc(i);
-
-		spin_lock_irqsave(&desc->lock, flags);
-		action = desc->action;
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
 		if (!action)
 			goto unlock;
 
@@ -68,7 +66,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_putc(p, '\n');
 	unlock:
-		spin_unlock_irqrestore(&desc->lock, flags);
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
 	}
 
 	return 0;
Index: linux-2.6/arch/cris/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/cris/kernel/irq.c
+++ linux-2.6/arch/cris/kernel/irq.c
@@ -99,7 +99,7 @@ asmlinkage void do_IRQ(int irq, struct p
 		printk("do_IRQ: stack overflow: %lX\n", sp);
 		show_stack(NULL, (unsigned long *)sp);
 	}
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
         irq_exit();
 	set_irq_regs(old_regs);
 }
Index: linux-2.6/arch/h8300/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/h8300/kernel/irq.c
+++ linux-2.6/arch/h8300/kernel/irq.c
@@ -176,7 +176,7 @@ void __init init_IRQ(void)
 asmlinkage void do_IRQ(int irq)
 {
 	irq_enter();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 }
 
Index: linux-2.6/arch/ia64/kernel/iosapic.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/iosapic.c
+++ linux-2.6/arch/ia64/kernel/iosapic.c
@@ -452,7 +452,7 @@ iosapic_ack_edge_irq (unsigned int irq)
 	irq_desc_t *idesc = irq_desc + irq;
 
 	irq_complete_move(irq);
-	move_native_irq(irq, idesc);
+	move_native_irq(irq);
 	/*
 	 * Once we have recorded IRQ_PENDING already, we can mask the
 	 * interrupt for real. This prevents IRQ storms from unhandled
Index: linux-2.6/arch/ia64/kernel/msi_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/msi_ia64.c
+++ linux-2.6/arch/ia64/kernel/msi_ia64.c
@@ -120,7 +120,7 @@ void ia64_teardown_msi_irq(unsigned int
 static void ia64_ack_msi_irq(unsigned int irq)
 {
 	irq_complete_move(irq);
-	move_native_irq(irq, &irq_desc[irq]);
+	move_native_irq(irq);
 	ia64_eoi();
 }
 
Index: linux-2.6/arch/ia64/sn/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/irq.c
+++ linux-2.6/arch/ia64/sn/kernel/irq.c
@@ -110,7 +110,7 @@ static void sn_ack_irq(unsigned int irq)
 	HUB_S((u64*)LOCAL_MMR_ADDR(SH_EVENT_OCCURRED_ALIAS), mask);
 	__set_bit(irq, (volatile void *)pda->sn_in_service_ivecs);
 
-	move_native_irq(irq, &irq_desc[irq]);
+	move_native_irq(irq);
 }
 
 static void sn_end_irq(unsigned int irq)
Index: linux-2.6/arch/ia64/sn/kernel/msi_sn.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/msi_sn.c
+++ linux-2.6/arch/ia64/sn/kernel/msi_sn.c
@@ -210,7 +210,7 @@ static void sn_set_msi_irq_affinity(unsi
 
 static void sn_ack_msi_irq(unsigned int irq)
 {
-	move_native_irq(irq, &irq_desc[irq]);
+	move_native_irq(irq);
 	ia64_eoi();
 }
 
Index: linux-2.6/arch/m32r/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m32r/kernel/irq.c
+++ linux-2.6/arch/m32r/kernel/irq.c
@@ -84,7 +84,7 @@ asmlinkage unsigned int do_IRQ(int irq,
 #ifdef CONFIG_DEBUG_STACKOVERFLOW
 	/* FIXME M32R */
 #endif
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 	set_irq_regs(old_regs);
 
Index: linux-2.6/arch/m68knommu/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m68knommu/kernel/irq.c
+++ linux-2.6/arch/m68knommu/kernel/irq.c
@@ -23,7 +23,7 @@ asmlinkage void do_IRQ(int irq, struct p
 	struct pt_regs *oldregs = set_irq_regs(regs);
 
 	irq_enter();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 
 	set_irq_regs(oldregs);
Index: linux-2.6/arch/parisc/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/parisc/kernel/irq.c
+++ linux-2.6/arch/parisc/kernel/irq.c
@@ -363,7 +363,7 @@ void do_cpu_irq_mask(struct pt_regs *reg
 		goto set_out;
 	}
 #endif
-	__do_IRQ(irq, irq_desc);
+	__do_IRQ(irq);
 
  out:
 	irq_exit();
Index: linux-2.6/arch/powerpc/platforms/powermac/pic.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/powermac/pic.c
+++ linux-2.6/arch/powerpc/platforms/powermac/pic.c
@@ -221,7 +221,7 @@ static irqreturn_t gatwick_action(int cp
 			continue;
 		irq += __ilog2(bits);
 		spin_unlock_irqrestore(&pmac_pic_lock, flags);
-		__do_IRQ(irq, &irq_desc[irq]);
+		__do_IRQ(irq);
 		spin_lock_irqsave(&pmac_pic_lock, flags);
 		rc = IRQ_HANDLED;
 	}
Index: linux-2.6/arch/um/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/um/kernel/irq.c
+++ linux-2.6/arch/um/kernel/irq.c
@@ -332,7 +332,7 @@ unsigned int do_IRQ(int irq, struct uml_
 {
 	struct pt_regs *old_regs = set_irq_regs((struct pt_regs *)regs);
 	irq_enter();
-	__do_IRQ(irq, &irq_desc[irq]);
+	__do_IRQ(irq);
 	irq_exit();
 	set_irq_regs(old_regs);
 	return 1;
Index: linux-2.6/drivers/parisc/dino.c
===================================================================
--- linux-2.6.orig/drivers/parisc/dino.c
+++ linux-2.6/drivers/parisc/dino.c
@@ -389,10 +389,9 @@ ilr_again:
 	do {
 		int local_irq = __ffs(mask);
 		int irq = dino_dev->global_irq[local_irq];
-		struct irq_desc *desc = to_irq_desc(irq);
 		DBG(KERN_DEBUG "%s(%d, %p) mask 0x%x\n",
 			__func__, irq, intr_dev, mask);
-		__do_IRQ(irq, desc);
+		__do_IRQ(irq);
 		mask &= ~(1 << local_irq);
 	} while (mask);
 
Index: linux-2.6/drivers/parisc/eisa.c
===================================================================
--- linux-2.6.orig/drivers/parisc/eisa.c
+++ linux-2.6/drivers/parisc/eisa.c
@@ -202,7 +202,6 @@ static irqreturn_t eisa_irq(int wax_irq,
 {
 	int irq = gsc_readb(0xfc01f000); /* EISA supports 16 irqs */
 	unsigned long flags;
-	struct irq_desc *desc;
         
 	spin_lock_irqsave(&eisa_irq_lock, flags);
 	/* read IRR command */
@@ -234,8 +233,7 @@ static irqreturn_t eisa_irq(int wax_irq,
 	}
 	spin_unlock_irqrestore(&eisa_irq_lock, flags);
 
-	desc = to_irq_desc(irq);
-	__do_IRQ(irq, desc);
+	__do_IRQ(irq);
    
 	spin_lock_irqsave(&eisa_irq_lock, flags);
 	/* unmask */
Index: linux-2.6/drivers/parisc/gsc.c
===================================================================
--- linux-2.6.orig/drivers/parisc/gsc.c
+++ linux-2.6/drivers/parisc/gsc.c
@@ -87,8 +87,7 @@ irqreturn_t gsc_asic_intr(int gsc_asic_i
 	do {
 		int local_irq = __ffs(irr);
 		unsigned int irq = gsc_asic->global_irq[local_irq];
-		struct irq_desc *desc = to_irq_desc(irq);
-		__do_IRQ(irq, desc);
+		__do_IRQ(irq);
 		irr &= ~(1 << local_irq);
 	} while (irr);
 
Index: linux-2.6/drivers/parisc/superio.c
===================================================================
--- linux-2.6.orig/drivers/parisc/superio.c
+++ linux-2.6/drivers/parisc/superio.c
@@ -99,7 +99,6 @@ superio_interrupt(int parent_irq, void *
 {
 	u8 results;
 	u8 local_irq;
-	struct irq_desc *desc;
 
 	/* Poll the 8259 to see if there's an interrupt. */
 	outb (OCW3_POLL,IC_PIC1+0);
@@ -140,8 +139,7 @@ superio_interrupt(int parent_irq, void *
 	}
 
 	/* Call the appropriate device's interrupt */
-	desc = to_irq_desc(local_irq);
-	__do_IRQ(local_irq, desc);
+	__do_IRQ(local_irq);
 
 	/* set EOI - forces a new interrupt if a lower priority device
 	 * still needs service.
Index: linux-2.6/drivers/xen/events.c
===================================================================
--- linux-2.6.orig/drivers/xen/events.c
+++ linux-2.6/drivers/xen/events.c
@@ -640,7 +640,7 @@ static void ack_dynirq(unsigned int irq)
 {
 	int evtchn = evtchn_from_irq(irq);
 
-	move_native_irq(irq, to_irq_desc(irq));
+	move_native_irq(irq);
 
 	if (VALID_EVTCHN(evtchn))
 		clear_evtchn(evtchn);
Index: linux-2.6/include/linux/irq.h
===================================================================
--- linux-2.6.orig/include/linux/irq.h
+++ linux-2.6/include/linux/irq.h
@@ -244,7 +244,7 @@ extern int setup_irq(unsigned int irq, s
 #ifdef CONFIG_GENERIC_PENDING_IRQ
 
 void set_pending_irq(unsigned int irq, cpumask_t mask);
-void move_native_irq(int irq, struct irq_desc *desc);
+void move_native_irq(int irq);
 void move_masked_irq(int irq, struct irq_desc *desc);
 
 #else /* CONFIG_GENERIC_PENDING_IRQ */
@@ -253,7 +253,7 @@ static inline void move_irq(int irq)
 {
 }
 
-static inline void move_native_irq(int irq, struct irq_desc *desc)
+static inline void move_native_irq(int irq)
 {
 }
 
@@ -269,7 +269,7 @@ static inline void set_pending_irq(unsig
 
 #else /* CONFIG_SMP */
 
-#define move_native_irq(x, y)
+#define move_native_irq(x)
 #define move_masked_irq(x, y)
 
 #endif /* CONFIG_SMP */
@@ -302,7 +302,7 @@ extern void handle_bad_irq(unsigned int
  * Monolithic do_IRQ implementation.
  */
 #ifndef CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ
-extern unsigned int __do_IRQ(unsigned int irq, struct irq_desc *desc);
+extern unsigned int __do_IRQ(unsigned int irq);
 #endif
 
 /*
@@ -319,7 +319,7 @@ static inline void generic_handle_irq_de
 	if (likely(desc->handle_irq))
 		desc->handle_irq(irq, desc);
 	else
-		__do_IRQ(irq, desc);
+		__do_IRQ(irq);
 #endif
 }
 
Index: linux-2.6/kernel/irq/handle.c
===================================================================
--- linux-2.6.orig/kernel/irq/handle.c
+++ linux-2.6/kernel/irq/handle.c
@@ -385,10 +385,11 @@ irqreturn_t handle_IRQ_event(unsigned in
  * This is the original x86 implementation which is used for every
  * interrupt type.
  */
-unsigned int __do_IRQ(unsigned int irq, struct irq_desc *desc)
+unsigned int __do_IRQ(unsigned int irq)
 {
 	struct irqaction *action;
 	unsigned int status;
+	struct irq_desc *desc = to_irq_desc(irq);
 
 	kstat_irqs_this_cpu(desc)++;
 	if (CHECK_IRQ_PER_CPU(desc->status)) {
Index: linux-2.6/kernel/irq/migration.c
===================================================================
--- linux-2.6.orig/kernel/irq/migration.c
+++ linux-2.6/kernel/irq/migration.c
@@ -57,8 +57,10 @@ void move_masked_irq(int irq, struct irq
 	cpus_clear(desc->pending_mask);
 }
 
-void move_native_irq(int irq, struct irq_desc *desc)
+void move_native_irq(int irq)
 {
+	struct irq_desc *desc = to_irq_desc(irq);
+
 	if (likely(!(desc->status & IRQ_MOVE_PENDING)))
 		return;
 
Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic.c
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -2426,10 +2426,8 @@ static void ack_x2apic_edge(unsigned int
 
 static void ack_apic_edge(unsigned int irq)
 {
-	struct irq_desc *desc = to_irq_desc(irq);
-
 	irq_complete_move(irq);
-	move_native_irq(irq, desc);
+	move_native_irq(irq);
 	ack_APIC_irq();
 }
 

[PATCH] irq: rename irq_desc() to to_irq_desc() - fix #2

[Yinghai Lu]

kstat_irqs related, so only CONFIG_DYN_ARRAY/SPARSE_IRQ will have kstat_irqs in irq_desc

other arch still have kstat_irqs in kstat

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/alpha/kernel/irq.c                 |    2 +-
 arch/alpha/kernel/irq_alpha.c           |    2 +-
 arch/arm/kernel/irq.c                   |    2 +-
 arch/arm/mach-ns9xxx/irq.c              |    3 ++-
 arch/avr32/kernel/irq.c                 |    4 ++--
 arch/cris/kernel/irq.c                  |    2 +-
 arch/frv/kernel/irq.c                   |    2 +-
 arch/h8300/kernel/irq.c                 |    2 +-
 arch/ia64/kernel/irq.c                  |    2 +-
 arch/ia64/kernel/irq_ia64.c             |    8 ++++----
 arch/m32r/kernel/irq.c                  |    2 +-
 arch/mips/kernel/irq.c                  |    2 +-
 arch/mips/sgi-ip22/ip22-int.c           |    2 +-
 arch/mips/sgi-ip22/ip22-time.c          |    2 +-
 arch/mips/sibyte/bcm1480/smp.c          |    2 +-
 arch/mips/sibyte/sb1250/smp.c           |    2 +-
 arch/mn10300/kernel/irq.c               |    2 +-
 arch/parisc/kernel/irq.c                |    2 +-
 arch/powerpc/kernel/irq.c               |    2 +-
 arch/powerpc/platforms/cell/interrupt.c |    2 +-
 arch/sh/kernel/irq.c                    |    2 +-
 arch/sparc64/kernel/irq.c               |    2 +-
 arch/sparc64/kernel/time.c              |    3 +--
 arch/um/kernel/irq.c                    |    2 +-
 arch/xtensa/kernel/irq.c                |    2 +-
 include/linux/irq.h                     |    5 +++++
 include/linux/kernel_stat.h             |    4 ++--
 kernel/irq/chip.c                       |   20 ++++++++++++++++++++
 kernel/irq/handle.c                     |   10 ++++++++++
 29 files changed, 67 insertions(+), 32 deletions(-)

Index: linux-2.6/arch/alpha/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq.c
+++ linux-2.6/arch/alpha/kernel/irq.c
@@ -89,7 +89,7 @@ show_interrupts(struct seq_file *p, void
 		seq_printf(p, "%10u ", kstat_irqs(irq));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(irq, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[irq]);
 #endif
 		seq_printf(p, " %14s", irq_desc[irq].chip->typename);
 		seq_printf(p, "  %c%s",
Index: linux-2.6/arch/alpha/kernel/irq_alpha.c
===================================================================
--- linux-2.6.orig/arch/alpha/kernel/irq_alpha.c
+++ linux-2.6/arch/alpha/kernel/irq_alpha.c
@@ -64,7 +64,7 @@ do_entInt(unsigned long type, unsigned l
 		smp_percpu_timer_interrupt(regs);
 		cpu = smp_processor_id();
 		if (cpu != boot_cpuid) {
-			irq_desc[RTC_IRQ].kstat_irqs[cpu]++;
+		        kstat_cpu(cpu).irqs[RTC_IRQ]++;
 		} else {
 			handle_irq(RTC_IRQ);
 		}
Index: linux-2.6/arch/arm/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/kernel/irq.c
+++ linux-2.6/arch/arm/kernel/irq.c
@@ -76,7 +76,7 @@ int show_interrupts(struct seq_file *p,
 
 		seq_printf(p, "%3d: ", i);
 		for_each_present_cpu(cpu)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
+			seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
 		seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
 		seq_printf(p, "  %s", action->name);
 		for (action = action->next; action; action = action->next)
Index: linux-2.6/arch/arm/mach-ns9xxx/irq.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-ns9xxx/irq.c
+++ linux-2.6/arch/arm/mach-ns9xxx/irq.c
@@ -63,6 +63,7 @@ static struct irq_chip ns9xxx_chip = {
 #else
 static void handle_prio_irq(unsigned int irq, struct irq_desc *desc)
 {
+	unsigned int cpu = smp_processor_id();
 	struct irqaction *action;
 	irqreturn_t action_ret;
 
@@ -71,7 +72,7 @@ static void handle_prio_irq(unsigned int
 	BUG_ON(desc->status & IRQ_INPROGRESS);
 
 	desc->status &= ~(IRQ_REPLAY | IRQ_WAITING);
-	kstat_irqs_this_cpu(desc)++;
+	kstat_cpu(cpu).irqs[irq]++;
 
 	action = desc->action;
 	if (unlikely(!action || (desc->status & IRQ_DISABLED)))
Index: linux-2.6/arch/avr32/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/avr32/kernel/irq.c
+++ linux-2.6/arch/avr32/kernel/irq.c
@@ -58,8 +58,8 @@ int show_interrupts(struct seq_file *p,
 
 		seq_printf(p, "%3d: ", i);
 		for_each_online_cpu(cpu)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
-		seq_printf(p, " %8s", desc->chip->name ? : "-");
+			seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
+		seq_printf(p, " %8s", irq_desc[i].chip->name ? : "-");
 		seq_printf(p, "  %s", action->name);
 		for (action = action->next; action; action = action->next)
 			seq_printf(p, ", %s", action->name);
Index: linux-2.6/arch/cris/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/cris/kernel/irq.c
+++ linux-2.6/arch/cris/kernel/irq.c
@@ -66,7 +66,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/frv/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/frv/kernel/irq.c
+++ linux-2.6/arch/frv/kernel/irq.c
@@ -74,7 +74,7 @@ int show_interrupts(struct seq_file *p,
 		if (action) {
 			seq_printf(p, "%3d: ", i);
 			for_each_present_cpu(cpu)
-				seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
+				seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
 			seq_printf(p, " %10s", irq_desc[i].chip->name ? : "-");
 			seq_printf(p, "  %s", action->name);
 			for (action = action->next;
Index: linux-2.6/arch/h8300/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/h8300/kernel/irq.c
+++ linux-2.6/arch/h8300/kernel/irq.c
@@ -196,7 +196,7 @@ int show_interrupts(struct seq_file *p,
 		if (!action)
 			goto unlock;
 		seq_printf(p, "%3d: ",i);
-		seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+		seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "-%-8s", irq_desc[i].name);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/ia64/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/irq.c
+++ linux-2.6/arch/ia64/kernel/irq.c
@@ -80,7 +80,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j) {
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 		}
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
Index: linux-2.6/arch/ia64/kernel/irq_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/irq_ia64.c
+++ linux-2.6/arch/ia64/kernel/irq_ia64.c
@@ -495,9 +495,9 @@ ia64_handle_irq (ia64_vector vector, str
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
 		if (unlikely(IS_LOCAL_TLB_FLUSH(vector))) {
 			smp_local_flush_tlb();
-			kstat_irqs_this_cpu(&irq_desc[vector])++;
+			kstat_this_cpu.irqs[vector]++;
 		} else if (unlikely(IS_RESCHEDULE(vector)))
-			kstat_irqs_this_cpu(&irq_desc[vector])++;
+			kstat_this_cpu.irqs[vector]++;
 		else {
 			int irq = local_vector_to_irq(vector);
 
@@ -553,9 +553,9 @@ void ia64_process_pending_intr(void)
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
 		if (unlikely(IS_LOCAL_TLB_FLUSH(vector))) {
 			smp_local_flush_tlb();
-			kstat_irqs_this_cpu(&irq_desc[vector])++;
+			kstat_this_cpu.irqs[vector]++;
 		} else if (unlikely(IS_RESCHEDULE(vector)))
-			kstat_irqs_this_cpu(&irq_desc[vector])++;
+			kstat_this_cpu.irqs[vector]++;
 		else {
 			struct pt_regs *old_regs = set_irq_regs(NULL);
 			int irq = local_vector_to_irq(vector);
Index: linux-2.6/arch/m32r/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/m32r/kernel/irq.c
+++ linux-2.6/arch/m32r/kernel/irq.c
@@ -52,7 +52,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/mips/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/mips/kernel/irq.c
+++ linux-2.6/arch/mips/kernel/irq.c
@@ -108,7 +108,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/mips/sgi-ip22/ip22-int.c
===================================================================
--- linux-2.6.orig/arch/mips/sgi-ip22/ip22-int.c
+++ linux-2.6/arch/mips/sgi-ip22/ip22-int.c
@@ -164,7 +164,7 @@ static void indy_buserror_irq(void)
 	int irq = SGI_BUSERR_IRQ;
 
 	irq_enter();
-	kstat_irqs_this_cpu(&irq_desc[irq])++;
+	kstat_this_cpu.irqs[irq]++;
 	ip22_be_interrupt(irq);
 	irq_exit();
 }
Index: linux-2.6/arch/mips/sgi-ip22/ip22-time.c
===================================================================
--- linux-2.6.orig/arch/mips/sgi-ip22/ip22-time.c
+++ linux-2.6/arch/mips/sgi-ip22/ip22-time.c
@@ -186,7 +186,7 @@ void indy_8254timer_irq(void)
 	char c;
 
 	irq_enter();
-	kstat_irqs_this_cpu(&irq_desc[irq])++;
+	kstat_this_cpu.irqs[irq]++;
 	printk(KERN_ALERT "Oops, got 8254 interrupt.\n");
 	ArcRead(0, &c, 1, &cnt);
 	ArcEnterInteractiveMode();
Index: linux-2.6/arch/mips/sibyte/bcm1480/smp.c
===================================================================
--- linux-2.6.orig/arch/mips/sibyte/bcm1480/smp.c
+++ linux-2.6/arch/mips/sibyte/bcm1480/smp.c
@@ -180,7 +180,7 @@ void bcm1480_mailbox_interrupt(void)
 	int cpu = smp_processor_id();
 	unsigned int action;
 
-	irq_desc[K_BCM1480_INT_MBOX_0_0].kstat_irqs[cpu]++;
+	kstat_this_cpu.irqs[K_BCM1480_INT_MBOX_0_0]++;
 	/* Load the mailbox register to figure out what we're supposed to do */
 	action = (__raw_readq(mailbox_0_regs[cpu]) >> 48) & 0xffff;
 
Index: linux-2.6/arch/mips/sibyte/sb1250/smp.c
===================================================================
--- linux-2.6.orig/arch/mips/sibyte/sb1250/smp.c
+++ linux-2.6/arch/mips/sibyte/sb1250/smp.c
@@ -168,7 +168,7 @@ void sb1250_mailbox_interrupt(void)
 	int cpu = smp_processor_id();
 	unsigned int action;
 
-	irq_desc[K_INT_MBOX_0].kstat_irqs[cpu]++;
+	kstat_this_cpu.irqs[K_INT_MBOX_0]++;
 	/* Load the mailbox register to figure out what we're supposed to do */
 	action = (____raw_readq(mailbox_regs[cpu]) >> 48) & 0xffff;
 
Index: linux-2.6/arch/mn10300/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/mn10300/kernel/irq.c
+++ linux-2.6/arch/mn10300/kernel/irq.c
@@ -201,7 +201,7 @@ int show_interrupts(struct seq_file *p,
 		if (action) {
 			seq_printf(p, "%3d: ", i);
 			for_each_present_cpu(cpu)
-				seq_printf(p, "%10u ", kstat_irqs_cpu(i, cpu));
+				seq_printf(p, "%10u ", kstat_cpu(cpu).irqs[i]);
 			seq_printf(p, " %14s.%u", irq_desc[i].chip->name,
 				   (GxICR(i) & GxICR_LEVEL) >>
 				   GxICR_LEVEL_SHIFT);
Index: linux-2.6/arch/parisc/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/parisc/kernel/irq.c
+++ linux-2.6/arch/parisc/kernel/irq.c
@@ -183,7 +183,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%3d: ", i);
 #ifdef CONFIG_SMP
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #else
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #endif
Index: linux-2.6/arch/powerpc/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/powerpc/kernel/irq.c
+++ linux-2.6/arch/powerpc/kernel/irq.c
@@ -190,7 +190,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%3d: ", i);
 #ifdef CONFIG_SMP
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #else
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #endif /* CONFIG_SMP */
Index: linux-2.6/arch/powerpc/platforms/cell/interrupt.c
===================================================================
--- linux-2.6.orig/arch/powerpc/platforms/cell/interrupt.c
+++ linux-2.6/arch/powerpc/platforms/cell/interrupt.c
@@ -254,7 +254,7 @@ static void handle_iic_irq(unsigned int
 		goto out_eoi;
 	}
 
-	kstat_irqs_this_cpu(desc)++;
+	kstat_cpu(cpu).irqs[irq]++;
 
 	/* Mark the IRQ currently in progress.*/
 	desc->status |= IRQ_INPROGRESS;
Index: linux-2.6/arch/sh/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/sh/kernel/irq.c
+++ linux-2.6/arch/sh/kernel/irq.c
@@ -51,7 +51,7 @@ int show_interrupts(struct seq_file *p,
 			goto unlock;
 		seq_printf(p, "%3d: ",i);
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 		seq_printf(p, " %14s", irq_desc[i].chip->name);
 		seq_printf(p, "-%-8s", irq_desc[i].name);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/sparc64/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/sparc64/kernel/irq.c
+++ linux-2.6/arch/sparc64/kernel/irq.c
@@ -185,7 +185,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #endif
 		seq_printf(p, " %9s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/sparc64/kernel/time.c
===================================================================
--- linux-2.6.orig/arch/sparc64/kernel/time.c
+++ linux-2.6/arch/sparc64/kernel/time.c
@@ -16,7 +16,6 @@
 #include <linux/param.h>
 #include <linux/string.h>
 #include <linux/mm.h>
-#include <linux/irq.h>
 #include <linux/interrupt.h>
 #include <linux/time.h>
 #include <linux/timex.h>
@@ -943,7 +942,7 @@ void timer_interrupt(int irq, struct pt_
 
 	irq_enter();
 
-	irq_desc[0].kstat_irqs[0]++;
+	kstat_this_cpu.irqs[0]++;
 
 	if (unlikely(!evt->event_handler)) {
 		printk(KERN_WARNING
Index: linux-2.6/arch/um/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/um/kernel/irq.c
+++ linux-2.6/arch/um/kernel/irq.c
@@ -42,7 +42,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/arch/xtensa/kernel/irq.c
===================================================================
--- linux-2.6.orig/arch/xtensa/kernel/irq.c
+++ linux-2.6/arch/xtensa/kernel/irq.c
@@ -99,7 +99,7 @@ int show_interrupts(struct seq_file *p,
 		seq_printf(p, "%10u ", kstat_irqs(i));
 #else
 		for_each_online_cpu(j)
-			seq_printf(p, "%10u ", kstat_irqs_cpu(i, j));
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
 #endif
 		seq_printf(p, " %14s", irq_desc[i].chip->typename);
 		seq_printf(p, "  %s", action->name);
Index: linux-2.6/include/linux/irq.h
===================================================================
--- linux-2.6.orig/include/linux/irq.h
+++ linux-2.6/include/linux/irq.h
@@ -219,8 +219,13 @@ extern struct irq_desc *irq_descX;
 
 #endif
 
+#ifdef CONFIG_HAVE_DYN_ARRAY
 #define kstat_irqs_this_cpu(DESC) \
 	((DESC)->kstat_irqs[smp_processor_id()])
+#else
+#define kstat_irqs_this_cpu(irq) \
+	kstat_this_cpu.irqs[irq]
+#endif
 
 /*
  * Migration helpers for obsolete names, they will go away:
Index: linux-2.6/include/linux/kernel_stat.h
===================================================================
--- linux-2.6.orig/include/linux/kernel_stat.h
+++ linux-2.6/include/linux/kernel_stat.h
@@ -28,7 +28,7 @@ struct cpu_usage_stat {
 
 struct kernel_stat {
 	struct cpu_usage_stat	cpustat;
-#ifndef CONFIG_GENRIC_HARDIRQS
+#ifndef CONFIG_HAVE_DYN_ARRAY
        unsigned int irqs[NR_IRQS];
 #endif
 };
@@ -41,7 +41,7 @@ DECLARE_PER_CPU(struct kernel_stat, ksta
 
 extern unsigned long long nr_context_switches(void);
 
-#ifndef CONFIG_GENERIC_HARDIRQS
+#ifndef CONFIG_HAVE_DYN_ARRAY
 static inline unsigned int kstat_irqs_cpu(unsigned int irq, int cpu)
 {
        return kstat_cpu(cpu).irqs[irq];
Index: linux-2.6/kernel/irq/chip.c
===================================================================
--- linux-2.6.orig/kernel/irq/chip.c
+++ linux-2.6/kernel/irq/chip.c
@@ -326,7 +326,11 @@ handle_simple_irq(unsigned int irq, stru
 	if (unlikely(desc->status & IRQ_INPROGRESS))
 		goto out_unlock;
 	desc->status &= ~(IRQ_REPLAY | IRQ_WAITING);
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 
 	action = desc->action;
 	if (unlikely(!action || (desc->status & IRQ_DISABLED)))
@@ -367,7 +371,11 @@ handle_level_irq(unsigned int irq, struc
 	if (unlikely(desc->status & IRQ_INPROGRESS))
 		goto out_unlock;
 	desc->status &= ~(IRQ_REPLAY | IRQ_WAITING);
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 
 	/*
 	 * If its disabled or no action available
@@ -414,7 +422,11 @@ handle_fasteoi_irq(unsigned int irq, str
 		goto out;
 
 	desc->status &= ~(IRQ_REPLAY | IRQ_WAITING);
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 
 	/*
 	 * If its disabled or no action available
@@ -479,7 +491,11 @@ handle_edge_irq(unsigned int irq, struct
 		goto out_unlock;
 	}
 
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 
 	/* Start handling the irq */
 	desc->chip->ack(irq);
@@ -534,7 +550,11 @@ handle_percpu_irq(unsigned int irq, stru
 {
 	irqreturn_t action_ret;
 
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 
 	if (desc->chip->ack)
 		desc->chip->ack(irq);
Index: linux-2.6/kernel/irq/handle.c
===================================================================
--- linux-2.6.orig/kernel/irq/handle.c
+++ linux-2.6/kernel/irq/handle.c
@@ -34,7 +34,11 @@ void
 handle_bad_irq(unsigned int irq, struct irq_desc *desc)
 {
 	print_irq_desc(irq, desc);
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 	ack_bad_irq(irq);
 }
 
@@ -391,7 +395,11 @@ unsigned int __do_IRQ(unsigned int irq)
 	unsigned int status;
 	struct irq_desc *desc = to_irq_desc(irq);
 
+#ifdef CONFIG_HAVE_DYN_ARRAY
 	kstat_irqs_this_cpu(desc)++;
+#else
+	kstat_irqs_this_cpu(irq)++;
+#endif
 	if (CHECK_IRQ_PER_CPU(desc->status)) {
 		irqreturn_t action_ret;
 
@@ -493,10 +501,12 @@ void early_init_irq_lock_class(void)
 }
 #endif
 
+#ifdef CONFIG_HAVE_DYN_ARRAY
 unsigned int kstat_irqs_cpu(unsigned int irq, int cpu)
 {
 	struct irq_desc *desc = to_irq_desc(irq);
 	return desc->kstat_irqs[cpu];
 }
+#endif
 EXPORT_SYMBOL(kstat_irqs_cpu);
 

[PATCH] irq: rename irq_desc() to to_irq_desc() - fix #3

[Yinghai Lu]

add create_irq_nr, and retore create_irq to old

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>

---
 arch/arm/mach-iop13xx/msi.c  |    4 ++--
 arch/ia64/kernel/iosapic.c   |    6 +++---
 arch/ia64/kernel/irq_ia64.c  |    2 +-
 arch/ia64/kernel/msi_ia64.c  |    2 +-
 arch/ia64/sn/kernel/msi_sn.c |    2 +-
 arch/x86/kernel/io_apic.c    |   18 +++++++++++++++---
 drivers/pci/htirq.c          |    2 +-
 drivers/pci/intel-iommu.c    |    2 +-
 include/linux/irq.h          |    3 ++-
 9 files changed, 27 insertions(+), 14 deletions(-)

Index: linux-2.6/arch/arm/mach-iop13xx/msi.c
===================================================================
--- linux-2.6.orig/arch/arm/mach-iop13xx/msi.c
+++ linux-2.6/arch/arm/mach-iop13xx/msi.c
@@ -125,7 +125,7 @@ void __init iop13xx_msi_init(void)
 /*
  * Dynamic irq allocate and deallocation
  */
-unsigned int create_irq(unsigned not_used)
+int create_irq(void)
 {
 	int irq, pos;
 
@@ -173,7 +173,7 @@ static struct irq_chip iop13xx_msi_chip
 
 int arch_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *desc)
 {
-	int id, irq = create_irq(0);
+	int id, irq = create_irq();
 	struct msi_msg msg;
 
 	if (irq < 0)
Index: linux-2.6/arch/ia64/kernel/iosapic.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/iosapic.c
+++ linux-2.6/arch/ia64/kernel/iosapic.c
@@ -532,7 +532,7 @@ iosapic_reassign_vector (int irq)
 	int new_irq;
 
 	if (iosapic_intr_info[irq].count) {
-		new_irq = create_irq(0);
+		new_irq = create_irq();
 		if (new_irq < 0)
 			panic("%s: out of interrupt vectors!\n", __func__);
 		printk(KERN_INFO "Reassigning vector %d to %d\n",
@@ -783,7 +783,7 @@ iosapic_register_intr (unsigned int gsi,
 			goto unlock_iosapic_lock;
 		}
 	} else
-		irq = create_irq(0);
+		irq = create_irq();
 
 	/* If vector is running out, we try to find a sharable vector */
 	if (irq < 0) {
@@ -922,7 +922,7 @@ iosapic_register_platform_intr (u32 int_
 		delivery = IOSAPIC_PMI;
 		break;
 	      case ACPI_INTERRUPT_INIT:
-		irq = create_irq(0);
+		irq = create_irq();
 		if (irq < 0)
 			panic("%s: out of interrupt vectors!\n", __func__);
 		vector = irq_to_vector(irq);
Index: linux-2.6/arch/ia64/kernel/irq_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/irq_ia64.c
+++ linux-2.6/arch/ia64/kernel/irq_ia64.c
@@ -403,7 +403,7 @@ void destroy_and_reserve_irq(unsigned in
 /*
  * Dynamic irq allocate and deallocation for MSI
  */
-unsigned int create_irq(unsigned int not_used)
+int create_irq(void)
 {
 	unsigned long flags;
 	int irq, vector, cpu;
Index: linux-2.6/arch/ia64/kernel/msi_ia64.c
===================================================================
--- linux-2.6.orig/arch/ia64/kernel/msi_ia64.c
+++ linux-2.6/arch/ia64/kernel/msi_ia64.c
@@ -84,7 +84,7 @@ int ia64_setup_msi_irq(struct pci_dev *p
 	int	irq, vector;
 	cpumask_t mask;
 
-	irq = create_irq(0);
+	irq = create_irq();
 	if (irq < 0)
 		return irq;
 
Index: linux-2.6/arch/ia64/sn/kernel/msi_sn.c
===================================================================
--- linux-2.6.orig/arch/ia64/sn/kernel/msi_sn.c
+++ linux-2.6/arch/ia64/sn/kernel/msi_sn.c
@@ -83,7 +83,7 @@ int sn_setup_msi_irq(struct pci_dev *pde
 	if (provider == NULL || provider->dma_map_consistent == NULL)
 		return -EINVAL;
 
-	irq = create_irq(0);
+	irq = create_irq();
 	if (irq < 0)
 		return irq;
 
Index: linux-2.6/arch/x86/kernel/io_apic.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/io_apic.c
+++ linux-2.6/arch/x86/kernel/io_apic.c
@@ -3039,7 +3039,7 @@ device_initcall(ioapic_init_sysfs);
 /*
  * Dynamic irq allocate and deallocation
  */
-unsigned int create_irq(unsigned int irq_want)
+unsigned int create_irq_nr(unsigned int irq_want)
 {
 	/* Allocate an unused irq */
 	unsigned int irq;
@@ -3074,6 +3074,18 @@ unsigned int create_irq(unsigned int irq
 	return irq;
 }
 
+int create_irq(void)
+{
+	int irq;
+
+	irq = create_irq_nr(nr_irqs - 1);
+
+	if (irq == 0)
+		irq = -1;
+
+	return irq;
+}
+
 void destroy_irq(unsigned int irq)
 {
 	unsigned long flags;
@@ -3346,7 +3358,7 @@ int arch_setup_msi_irq(struct pci_dev *d
 
 	irq_want = build_irq_for_pci_dev(dev) + 0x100;
 
-	irq = create_irq(irq_want);
+	irq = create_irq_nr(irq_want);
 	if (irq == 0)
 		return -1;
 
@@ -3388,7 +3400,7 @@ int arch_setup_msi_irqs(struct pci_dev *
 	irq_want = build_irq_for_pci_dev(dev) + 0x100;
 	sub_handle = 0;
 	list_for_each_entry(desc, &dev->msi_list, list) {
-		irq = create_irq(irq_want--);
+		irq = create_irq_nr(irq_want--);
 		if (irq == 0)
 			return -1;
 #ifdef CONFIG_INTR_REMAP
Index: linux-2.6/drivers/pci/htirq.c
===================================================================
--- linux-2.6.orig/drivers/pci/htirq.c
+++ linux-2.6/drivers/pci/htirq.c
@@ -139,7 +139,7 @@ int __ht_create_irq(struct pci_dev *dev,
 	cfg->msg.address_hi = 0xffffffff;
 
 	irq_want= build_irq_for_pci_dev(dev);
-	irq = create_irq(irq_want + idx);
+	irq = create_irq_nr(irq_want + idx);
 	if (irq == 0) {
 		kfree(cfg);
 		return -EBUSY;
Index: linux-2.6/drivers/pci/intel-iommu.c
===================================================================
--- linux-2.6.orig/drivers/pci/intel-iommu.c
+++ linux-2.6/drivers/pci/intel-iommu.c
@@ -915,7 +915,7 @@ int dmar_set_interrupt(struct intel_iomm
 {
 	int irq, ret;
 
-	irq = create_irq(0);
+	irq = create_irq();
 	if (!irq) {
 		printk(KERN_ERR "IOMMU: no free vectors\n");
 		return -EINVAL;
Index: linux-2.6/include/linux/irq.h
===================================================================
--- linux-2.6.orig/include/linux/irq.h
+++ linux-2.6/include/linux/irq.h
@@ -396,7 +396,8 @@ extern void set_irq_noprobe(unsigned int
 extern void set_irq_probe(unsigned int irq);
 
 /* Handle dynamic irq creation and destruction */
-extern unsigned int create_irq(unsigned int irq_want);
+extern unsigned int create_irq_nr(unsigned int irq_want);
+extern int create_irq(void);
 extern void destroy_irq(unsigned int irq);
 
 /* Test to see if a driver has successfully requested an irq */

Re: [PATCH] irq: rename irq_desc() to to_irq_desc() - fix #3

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> add create_irq_nr, and retore create_irq to old

applied all 3 patches to tip/irq/sparseirq - thanks Yinghai!

	Ingo

Re: [PATCH] irq: rename irq_desc() to to_irq_desc() - fix #3

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
> 
> > add create_irq_nr, and retore create_irq to old
> 
> applied all 3 patches to tip/irq/sparseirq - thanks Yinghai!

build fix for tip/master, from tip/x86/spinlocks. I merged that branch 
into tip/irq/sparseirq and applied this fix.

	Ingo

------------>
>From 1f73df742c77472d2e14e47b8c568218952af1d7 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Tue, 19 Aug 2008 02:33:39 +0200
Subject: [PATCH] irq: build fix

fix:

 arch/x86/xen/spinlock.c: In function 'xen_spin_lock_slow':
 arch/x86/xen/spinlock.c:90: error: 'struct kernel_stat' has no member named 'irqs'

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/xen/spinlock.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/xen/spinlock.c b/arch/x86/xen/spinlock.c
index 8dc4d31..dc21378 100644
--- a/arch/x86/xen/spinlock.c
+++ b/arch/x86/xen/spinlock.c
@@ -87,7 +87,7 @@ static noinline int xen_spin_lock_slow(struct raw_spinlock *lock)
 
 	/* block until irq becomes pending */
 	xen_poll_irq(irq);
-	kstat_this_cpu.irqs[irq]++;
+	kstat_irqs_this_cpu(&irq_desc[irq])++;
 
 out:
 	unspinning_lock(xl);

Re: [PATCH] irq: rename irq_desc() to to_irq_desc() - fix #3

[Yinghai Lu]

On Mon, Aug 18, 2008 at 5:38 PM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Ingo Molnar <mingo@elte.hu> wrote:
>
>> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>>
>> > add create_irq_nr, and retore create_irq to old
>>
>> applied all 3 patches to tip/irq/sparseirq - thanks Yinghai!
>
> build fix for tip/master, from tip/x86/spinlocks. I merged that branch
> into tip/irq/sparseirq and applied this fix.
>
>        Ingo
>
> ------------>
> From 1f73df742c77472d2e14e47b8c568218952af1d7 Mon Sep 17 00:00:00 2001
> From: Ingo Molnar <mingo@elte.hu>
> Date: Tue, 19 Aug 2008 02:33:39 +0200
> Subject: [PATCH] irq: build fix
>
> fix:
>
>  arch/x86/xen/spinlock.c: In function 'xen_spin_lock_slow':
>  arch/x86/xen/spinlock.c:90: error: 'struct kernel_stat' has no member named 'irqs'
>
> Signed-off-by: Ingo Molnar <mingo@elte.hu>
> ---
>  arch/x86/xen/spinlock.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/arch/x86/xen/spinlock.c b/arch/x86/xen/spinlock.c
> index 8dc4d31..dc21378 100644
> --- a/arch/x86/xen/spinlock.c
> +++ b/arch/x86/xen/spinlock.c
> @@ -87,7 +87,7 @@ static noinline int xen_spin_lock_slow(struct raw_spinlock *lock)
>
>        /* block until irq becomes pending */
>        xen_poll_irq(irq);
> -       kstat_this_cpu.irqs[irq]++;
> +       kstat_irqs_this_cpu(&irq_desc[irq])++;
>
>  out:
>        unspinning_lock(xl);
>

should be

kstat_irqs_this_cpu(to_irq_desc(irq))++;

otherwise when CONFIG_HAVE_SPARSE_IRQ is set, it will fail.

when sparse_irq is set, only have irq_descX list...

YH

Re: [PATCH] irq: rename irq_desc() to to_irq_desc() - fix #3

[Ingo Molnar]

* Yinghai Lu <yhlu.kernel@gmail.com> wrote:

> > -       kstat_this_cpu.irqs[irq]++;
> > +       kstat_irqs_this_cpu(&irq_desc[irq])++;
> >
> >  out:
> >        unspinning_lock(xl);
> >
> 
> should be
> 
> kstat_irqs_this_cpu(to_irq_desc(irq))++;
> 
> otherwise when CONFIG_HAVE_SPARSE_IRQ is set, it will fail.
> 
> when sparse_irq is set, only have irq_descX list...

ok, fixed it - updated patch below. Build testing just triggered it as 
well.

	Ingo

----------->
>From ba0540e8ae86abbe55b643e230ce276984fbb650 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Tue, 19 Aug 2008 02:33:39 +0200
Subject: [PATCH] irq: build fix

fix:

 arch/x86/xen/spinlock.c: In function 'xen_spin_lock_slow':
 arch/x86/xen/spinlock.c:90: error: 'struct kernel_stat' has no member named 'irqs'

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/xen/spinlock.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/xen/spinlock.c b/arch/x86/xen/spinlock.c
index 8dc4d31..bfb1707 100644
--- a/arch/x86/xen/spinlock.c
+++ b/arch/x86/xen/spinlock.c
@@ -87,7 +87,7 @@ static noinline int xen_spin_lock_slow(struct raw_spinlock *lock)
 
 	/* block until irq becomes pending */
 	xen_poll_irq(irq);
-	kstat_this_cpu.irqs[irq]++;
+	kstat_irqs_this_cpu(to_irq_desc(irq))++;
 
 out:
 	unspinning_lock(xl);

[PATCH] OpenRD: Enable SD/UART selection for serial port 1

[Tanmay Upadhyay]

This patch enables user to use serial port 1 of the OpenRD device for SDIO
or UART(RS232/RS485). The selection can be done through kernel parameter.

By default the port would be used for SDIO. To select RS232 or RS485 mode,
pass string "uart=232" or "uart=485" respectively in the kernel parameters.
"uart=485" is ignored on OpenRD-Base as it doesn't have RS485 port.

Signed-off-by: Tanmay Upadhyay <tanmay.upadhyay@einfochips.com>
---
 arch/arm/mach-kirkwood/openrd-setup.c |   67 ++++++++++++++++++++++++++++++++-
 1 files changed, 66 insertions(+), 1 deletions(-)

diff --git a/arch/arm/mach-kirkwood/openrd-setup.c b/arch/arm/mach-kirkwood/openrd-setup.c
index ad3f1ec..fb4bfad 100644
--- a/arch/arm/mach-kirkwood/openrd-setup.c
+++ b/arch/arm/mach-kirkwood/openrd-setup.c
@@ -15,6 +15,8 @@
 #include <linux/mtd/partitions.h>
 #include <linux/ata_platform.h>
 #include <linux/mv643xx_eth.h>
+#include <linux/io.h>
+#include <linux/gpio.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <mach/kirkwood.h>
@@ -56,16 +58,58 @@ static struct mvsdio_platform_data openrd_mvsdio_data = {
 };
 
 static unsigned int openrd_mpp_config[] __initdata = {
+	MPP12_SD_CLK,
+	MPP13_SD_CMD,
+	MPP14_SD_D0,
+	MPP15_SD_D1,
+	MPP16_SD_D2,
+	MPP17_SD_D3,
 	MPP29_GPIO,
 	0
 };
 
+static int sd_uart_selection(void)
+{
+	/* Parse boot_command_line string uart=no/232/485 */
+	char *ptr = strstr(boot_command_line, "uart=");
+
+	/* Default is SD. Change if required, for UART */
+	if (ptr != NULL) {
+		if (!strncmp(ptr + 5, "232", 3)) {
+			/* Configure MPP for UART */
+			openrd_mpp_config[1] = MPP13_UART1_TXD;
+			openrd_mpp_config[2] = MPP14_UART1_RXD;
+
+			return 232;
+		} else if (!strncmp(ptr + 5, "485", 3)) {
+			/* OpenRD-Base doesn't have RS485. Treat is as an
+			 * unknown argument & just have default setting -
+			 * which is SD */
+			if (machine_is_openrd_base())
+				return 0;
+
+			/* Configure MPP for UART */
+			openrd_mpp_config[1] = MPP13_UART1_TXD;
+			openrd_mpp_config[2] = MPP14_UART1_RXD;
+
+			return 485;
+		}
+	}
+	return 0;
+}
+
 static void __init openrd_init(void)
 {
+	int uart1;
+
 	/*
 	 * Basic setup. Needs to be called early.
 	 */
 	kirkwood_init();
+
+	/* This function modifies MPP config according to boot argument */
+	uart1 = sd_uart_selection();
+
 	kirkwood_mpp_conf(openrd_mpp_config);
 
 	kirkwood_uart0_init();
@@ -77,9 +121,30 @@ static void __init openrd_init(void)
 	if (machine_is_openrd_client())
 		kirkwood_ge01_init(&openrd_ge01_data);
 	kirkwood_sata_init(&openrd_sata_data);
-	kirkwood_sdio_init(&openrd_mvsdio_data);
 
 	kirkwood_i2c_init();
+
+	if (!uart1) {
+		/* Select SD
+		 * Pin # 34: 0 => UART1, 1 => SD */
+		writel(readl(GPIO_OUT(34)) | 4, GPIO_OUT(34));
+
+		kirkwood_sdio_init(&openrd_mvsdio_data);
+	} else {
+		/* Select UART1
+		 * Pin # 34: 0 => UART1, 1 => SD */
+		writel(readl(GPIO_OUT(34)) & ~(4), GPIO_OUT(34));
+
+		/* Select RS232 OR RS485
+		 * Pin # 28: 0 => RS232, 1 => RS485 */
+		if (uart1 == 232)
+			writel(readl(GPIO_OUT(28)) & ~(0x10000000),
+				GPIO_OUT(28));
+		else
+			writel(readl(GPIO_OUT(28)) | 0x10000000, GPIO_OUT(28));
+
+		kirkwood_uart1_init();
+	}
 }
 
 static int __init openrd_pci_init(void)
-- 
1.6.6.1

-- 
_____________________________________________________________________
Disclaimer: This e-mail message and all attachments transmitted with it
are intended solely for the use of the addressee and may contain legally
privileged and confidential information. If the reader of this message
is not the intended recipient, or an employee or agent responsible for
delivering this message to the intended recipient, you are hereby
notified that any dissemination, distribution, copying, or other use of
this message or its attachments is strictly prohibited. If you have
received this message in error, please notify the sender immediately by
replying to this message and please delete it from your computer. Any
views expressed in this message are those of the individual sender
unless otherwise stated.Company has taken enough precautions to prevent
the spread of viruses. However the company accepts no liability for any
damage caused by any virus transmitted by this email.
_____________________________________________________________________
 

[RFC PATCH 1/5] x86/Kconfig: Add Page Cache Accounting entry

[Liu Yuan]

From: Liu Yuan <tailai.ly@taobao.com>

Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
---
 arch/x86/Kconfig.debug |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index 615e188..f29e32d 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -304,4 +304,13 @@ config DEBUG_STRICT_USER_COPY_CHECKS
 
 	  If unsure, or if you run an older (pre 4.4) gcc, say N.
 
+config PAGE_CACHE_ACCT
+	bool "Page cache accounting"
+	---help---
+	  Enabling this options to account for page cache hit/missed number of
+	  times. This would allow user space applications get better knowledge
+	  of underlying page cache system by reading virtual file. The statitics
+	  per partition are collected.
+
+	  If unsure, say N.
 endmenu
-- 
1.7.0.4

[RFC PATCH 2/5] block: Add functions and data types for Page Cache Accounting

[Liu Yuan]

From: Liu Yuan <tailai.ly@taobao.com>

These functions and data types are based on the percpu
disk stats infrastructure.

Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
---
 include/linux/genhd.h |   56 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 56 insertions(+), 0 deletions(-)

diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index c0d5f69..4f0257c 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -86,6 +86,11 @@ struct disk_stats {
 	unsigned long ticks[2];
 	unsigned long io_ticks;
 	unsigned long time_in_queue;
+#ifdef CONFIG_PAGE_CACHE_ACCT
+	unsigned long page_cache_readpages;
+	unsigned long page_cache_hit[2];
+	unsigned long page_cache_missed[2];
+#endif
 };
 
 #define PARTITION_META_INFO_VOLNAMELTH	64
@@ -400,6 +405,57 @@ static inline void free_part_info(struct hd_struct *part)
 	kfree(part->info);
 }
 
+#ifdef CONFIG_PAGE_CACHE_ACCT
+static inline void page_cache_acct_readpages(struct super_block *sb, int nr_pages)
+{
+	struct block_device *bdev = sb->s_bdev;
+	struct hd_struct *part;
+	int cpu;
+	if (likely(bdev) && likely(part = bdev->bd_part)) {
+		cpu = part_stat_lock();
+		part_stat_add(cpu, part, page_cache_readpages, nr_pages);
+		part_stat_unlock();
+	}
+}
+static inline void page_cache_acct_hit(struct super_block *sb, int rw)
+{
+	struct block_device *bdev = sb->s_bdev;
+	struct hd_struct *part;
+	int cpu;
+	if (likely(bdev) && likely(part = bdev->bd_part)) {
+		cpu = part_stat_lock();
+		part_stat_inc(cpu, part, page_cache_hit[rw]);
+		part_stat_unlock();
+	}
+}
+
+static inline void page_cache_acct_missed(struct super_block *sb, int rw)
+{
+	struct block_device *bdev = sb->s_bdev;
+	struct hd_struct *part;
+	int cpu;
+	if (likely(bdev) && likely(part = bdev->bd_part)) {
+		cpu = part_stat_lock();
+		part_stat_inc(cpu, part, page_cache_missed[rw]);
+		part_stat_unlock();
+	}
+}
+
+#else /* !CONFIG_PAGE_CACHE_ACCT */
+static inline void page_cache_acct_readpages(struct super_block *sb, int nr_pages)
+{
+}
+
+static inline void page_cache_acct_hit(struct super_block *sb, int rw)
+{
+}
+
+static inline void page_cache_acct_missed(struct super_block *sb, int rw)
+{
+}
+
+#endif /* CONFIG_PAGE_CACHE_ACCT */
+
 /* block/blk-core.c */
 extern void part_round_stats(int cpu, struct hd_struct *part);
 
-- 
1.7.0.4

[RFC PATCH 3/5] block: Make Page Cache counters work with sysfs

[Liu Yuan]

From: Liu Yuan <tailai.ly@taobao.com>

Three counters are exported to the userspace from
/sys/block/sdx/{,sdxx}/page_cache_stats.

Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
---
 block/genhd.c         |    6 ++++++
 fs/partitions/check.c |   23 +++++++++++++++++++++++
 include/linux/genhd.h |    4 ++++
 3 files changed, 33 insertions(+), 0 deletions(-)

diff --git a/block/genhd.c b/block/genhd.c
index cbf1112..af9e7f8 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -988,6 +988,9 @@ static struct device_attribute dev_attr_fail_timeout =
 	__ATTR(io-timeout-fail,  S_IRUGO|S_IWUSR, part_timeout_show,
 		part_timeout_store);
 #endif
+#ifdef CONFIG_PAGE_CACHE_ACCT
+static DEVICE_ATTR(page_cache_stats, S_IRUGO, part_page_cache_stats_show, NULL);
+#endif
 
 static struct attribute *disk_attrs[] = {
 	&dev_attr_range.attr,
@@ -1006,6 +1009,9 @@ static struct attribute *disk_attrs[] = {
 #ifdef CONFIG_FAIL_IO_TIMEOUT
 	&dev_attr_fail_timeout.attr,
 #endif
+#ifdef CONFIG_PAGE_CACHE_ACCT
+	&dev_attr_page_cache_stats.attr,
+#endif
 	NULL
 };
 
diff --git a/fs/partitions/check.c b/fs/partitions/check.c
index 9c21119..e882e95 100644
--- a/fs/partitions/check.c
+++ b/fs/partitions/check.c
@@ -316,6 +316,23 @@ ssize_t part_fail_store(struct device *dev,
 }
 #endif
 
+#ifdef CONFIG_PAGE_CACHE_ACCT
+ssize_t part_page_cache_stats_show(struct device *dev,
+					  struct device_attribute * attr,
+					  char *buf)
+{
+	struct hd_struct  *p = dev_to_part(dev);
+
+	return sprintf(buf,
+			"%8lu %8lu %8lu %8lu %8lu\n ",
+			part_stat_read(p, page_cache_readpages),
+			part_stat_read(p, page_cache_missed[READ]),
+			part_stat_read(p, page_cache_hit[READ]),
+			part_stat_read(p, page_cache_missed[WRITE]),
+			part_stat_read(p, page_cache_hit[WRITE]));
+}
+#endif
+
 static DEVICE_ATTR(partition, S_IRUGO, part_partition_show, NULL);
 static DEVICE_ATTR(start, S_IRUGO, part_start_show, NULL);
 static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL);
@@ -329,6 +346,9 @@ static DEVICE_ATTR(inflight, S_IRUGO, part_inflight_show, NULL);
 static struct device_attribute dev_attr_fail =
 	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, part_fail_show, part_fail_store);
 #endif
+#ifdef CONFIG_PAGE_CACHE_ACCT
+static DEVICE_ATTR(page_cache_stats, S_IRUGO, part_page_cache_stats_show, NULL);
+#endif
 
 static struct attribute *part_attrs[] = {
 	&dev_attr_partition.attr,
@@ -342,6 +362,9 @@ static struct attribute *part_attrs[] = {
 #ifdef CONFIG_FAIL_MAKE_REQUEST
 	&dev_attr_fail.attr,
 #endif
+#ifdef CONFIG_PAGE_CACHE_ACCT
+	&dev_attr_page_cache_stats.attr,
+#endif
 	NULL
 };
 
diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index 4f0257c..0ecd165 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -682,6 +682,10 @@ extern ssize_t part_fail_store(struct device *dev,
 			       struct device_attribute *attr,
 			       const char *buf, size_t count);
 #endif /* CONFIG_FAIL_MAKE_REQUEST */
+#ifdef CONFIG_PAGE_CACHE_ACCT
+extern ssize_t part_page_cache_stats_show(struct device *dev,
+					  struct device_attribute *attr, char *buf);
+#endif /*  CONFIG_PAGE_CACHE_ACCT */
 
 static inline void hd_ref_init(struct hd_struct *part)
 {
-- 
1.7.0.4

[RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Liu Yuan]

From: Liu Yuan <tailai.ly@taobao.com>

Hit/Miss accountings are request-centric: that is, single request
would either cause one hit or one miss to be accounted for the very
first time that kernel query the page cache. In some rare error
conditions, kernel would re-query the page cache, but we donnot
account for it and ignore it for simplicity.

Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
---
 mm/filemap.c |   26 ++++++++++++++++++++++----
 1 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/mm/filemap.c b/mm/filemap.c
index 83a45d3..5388b2a 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -1009,10 +1009,15 @@ static void do_generic_file_read(struct file *filp, loff_t *ppos,
 		pgoff_t end_index;
 		loff_t isize;
 		unsigned long nr, ret;
+		int retry_find = 0;
 
 		cond_resched();
 find_page:
 		page = find_get_page(mapping, index);
+		if (likely(!retry_find) && page && PageUptodate(page))
+			page_cache_acct_hit(inode->i_sb, READ);
+		else
+			page_cache_acct_missed(inode->i_sb, READ);
 		if (!page) {
 			page_cache_sync_readahead(mapping,
 					ra, filp,
@@ -1137,6 +1142,7 @@ readpage:
 		if (unlikely(error)) {
 			if (error == AOP_TRUNCATED_PAGE) {
 				page_cache_release(page);
+				retry_find = 1;
 				goto find_page;
 			}
 			goto readpage_error;
@@ -1153,6 +1159,7 @@ readpage:
 					 */
 					unlock_page(page);
 					page_cache_release(page);
+					retry_find = 1;
 					goto find_page;
 				}
 				unlock_page(page);
@@ -1185,8 +1192,10 @@ no_cached_page:
 						index, GFP_KERNEL);
 		if (error) {
 			page_cache_release(page);
-			if (error == -EEXIST)
+			if (error == -EEXIST) {
+				retry_find = 1;
 				goto find_page;
+			}
 			desc->error = error;
 			goto out;
 		}
@@ -1543,6 +1552,7 @@ int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	struct page *page;
 	pgoff_t size;
 	int ret = 0;
+	int rw = !!(vmf->flags & FAULT_FLAG_WRITE);
 
 	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
 	if (offset >= size)
@@ -1552,6 +1562,10 @@ int filemap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	 * Do we have something in the page cache already?
 	 */
 	page = find_get_page(mapping, offset);
+	if (page && page->mapping && PageUptodate(page))
+		page_cache_acct_hit(inode->i_sb, rw);
+	else
+		page_cache_acct_missed(inode->i_sb, rw);
 	if (likely(page)) {
 		/*
 		 * We found the page, so try async readahead before
@@ -2227,20 +2241,24 @@ struct page *grab_cache_page_write_begin(struct address_space *mapping,
 		gfp_notmask = __GFP_FS;
 repeat:
 	page = find_lock_page(mapping, index);
-	if (page)
+	if (page) {
+		page_cache_acct_hit(mapping->host->i_sb, WRITE);
 		return page;
+	}
 
 	page = __page_cache_alloc(mapping_gfp_mask(mapping) & ~gfp_notmask);
 	if (!page)
-		return NULL;
+		goto out;
 	status = add_to_page_cache_lru(page, mapping, index,
 						GFP_KERNEL & ~gfp_notmask);
 	if (unlikely(status)) {
 		page_cache_release(page);
 		if (status == -EEXIST)
 			goto repeat;
-		return NULL;
+		page = NULL;
 	}
+out:
+	page_cache_acct_missed(mapping->host->i_sb, WRITE);
 	return page;
 }
 EXPORT_SYMBOL(grab_cache_page_write_begin);
-- 
1.7.0.4

[RFC PATCH 5/5] mm: Add readpages accounting

[Liu Yuan]

From: Liu Yuan <tailai.ly@taobao.com>

The _readpages_ counter simply counts how many pages the kernel
really request from the disk, either by readahead module or
aop->readpage() when readahead window equals 0.

This counter is request-centric and doesnot check read errors
since the read requests are issued to the block layer already.

Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
---
 mm/filemap.c   |    1 +
 mm/readahead.c |    2 ++
 2 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/mm/filemap.c b/mm/filemap.c
index 5388b2a..d638391 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -1137,6 +1137,7 @@ readpage:
 		 */
 		ClearPageError(page);
 		/* Start the actual read. The read will unlock the page. */
+		page_cache_acct_readpages(mapping->host->i_sb, 1);
 		error = mapping->a_ops->readpage(filp, page);
 
 		if (unlikely(error)) {
diff --git a/mm/readahead.c b/mm/readahead.c
index 77506a2..483acb8 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -112,6 +112,8 @@ static int read_pages(struct address_space *mapping, struct file *filp,
 	unsigned page_idx;
 	int ret;
 
+	page_cache_acct_readpages(mapping->host->i_sb, nr_pages);
+
 	if (mapping->a_ops->readpages) {
 		ret = mapping->a_ops->readpages(filp, mapping, pages, nr_pages);
 		/* Clean up the remaining pages */
-- 
1.7.0.4

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Ingo Molnar]

* Liu Yuan <namei.unix@gmail.com> wrote:

> +		if (likely(!retry_find) && page && PageUptodate(page))
> +			page_cache_acct_hit(inode->i_sb, READ);
> +		else
> +			page_cache_acct_missed(inode->i_sb, READ);

Sigh.

This would make such a nice tracepoint or sw perf event. It could be collected in a 
'count' form, equivalent to the stats you are aiming for here, or it could even be 
traced, if someone is interested in such details.

It could be mixed with other events, enriching multiple apps at once.

But, instead of trying to improve those aspects of our existing instrumentation 
frameworks, mm/* is gradually growing its own special instrumentation hacks, missing 
the big picture and fragmenting the instrumentation space some more.

That trend is somewhat sad.

Thanks,

	Ingo

Re: [RFC PATCH 1/5] x86/Kconfig: Add Page Cache Accounting entry

[Randy Dunlap]

On Wed,  2 Mar 2011 16:38:06 +0800 Liu Yuan wrote:

> From: Liu Yuan <tailai.ly@taobao.com>
> 
> Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
> ---
>  arch/x86/Kconfig.debug |    9 +++++++++
>  1 files changed, 9 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
> index 615e188..f29e32d 100644
> --- a/arch/x86/Kconfig.debug
> +++ b/arch/x86/Kconfig.debug
> @@ -304,4 +304,13 @@ config DEBUG_STRICT_USER_COPY_CHECKS
>  
>  	  If unsure, or if you run an older (pre 4.4) gcc, say N.
>  
> +config PAGE_CACHE_ACCT
> +	bool "Page cache accounting"
> +	---help---
> +	  Enabling this options to account for page cache hit/missed number of
> +	  times. This would allow user space applications get better knowledge
> +	  of underlying page cache system by reading virtual file. The statitics
> +	  per partition are collected.
> +
> +	  If unsure, say N.
>  endmenu
> -- 

rewrite:

	  Enable this option to provide for page cache hit/miss counters.
	  This allows userspace applications to obtain better knowledge of the
	  underlying page cache subsystem by reading a virtual file.
	  Statistics are collect per partition.

questions:
	what virtual file?
	what kind of partition?

---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Dave Hansen]

On Wed, 2011-03-02 at 09:45 +0100, Ingo Molnar wrote:
> But, instead of trying to improve those aspects of our existing instrumentation 
> frameworks, mm/* is gradually growing its own special instrumentation hacks, missing 
> the big picture and fragmenting the instrumentation space some more.
> 
> That trend is somewhat sad. 

Go any handy examples of how you'd like to see these done?

We're trying to add a batch of these for transparent huge pages, and
there was a similar set for KSM, so there's certainly no shortage of
potential sites.

-- Dave

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Ingo Molnar]

* Dave Hansen <dave@linux.vnet.ibm.com> wrote:

> On Wed, 2011-03-02 at 09:45 +0100, Ingo Molnar wrote:
> > But, instead of trying to improve those aspects of our existing instrumentation 
> > frameworks, mm/* is gradually growing its own special instrumentation hacks, missing 
> > the big picture and fragmenting the instrumentation space some more.
> > 
> > That trend is somewhat sad. 
> 
> Go any handy examples of how you'd like to see these done?

There's a very, very old branch in tip:tracing/mm (by Steve) that shows off some of 
the concepts that could be introduced, to 'dump' current MM state via an extension 
to the tracepoints APIs:

3383e37ea796: tracing, page-allocator: Add a postprocessing script for page-allocator-related ftrace events
c33b3596bc38: tracing, page-allocator: Add trace event for page traffic related to the buddy lists
0d524fb734bc: tracing, mm: Add trace events for anti-fragmentation falling back to other migratetypes
b9a28177eedf: tracing, page-allocator: Add trace events for page allocation and page freeing
807243eb20b2: Merge branch 'perfcounters/urgent' into tracing/mm
08b6cb88eeb5: perf_counter tools: Provide default bfd_demangle() function in case it's not around
eb4671011887: tracing/mm: rename 'trigger' file to 'dump_range'
1487a7a1ff99: tracing/mm: fix mapcount trace record field
dcac8cdac1d4: tracing/mm: add page frame snapshot trace

That's just a demo in essence - showing what things could be done in this area.

You can pick those commits up via running:

  http://people.redhat.com/mingo/tip.git/README

Thanks,

	Ingo

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Wu Fengguang]

On Thu, Mar 03, 2011 at 02:49:53AM +0800, Ingo Molnar wrote:
> 
> * Dave Hansen <dave@linux.vnet.ibm.com> wrote:
> 
> > On Wed, 2011-03-02 at 09:45 +0100, Ingo Molnar wrote:
> > > But, instead of trying to improve those aspects of our existing instrumentation 
> > > frameworks, mm/* is gradually growing its own special instrumentation hacks, missing 
> > > the big picture and fragmenting the instrumentation space some more.
> > > 
> > > That trend is somewhat sad. 
> > 
> > Go any handy examples of how you'd like to see these done?
> 
> There's a very, very old branch in tip:tracing/mm (by Steve) that shows off some of 
> the concepts that could be introduced, to 'dump' current MM state via an extension 
> to the tracepoints APIs:
> 
> 3383e37ea796: tracing, page-allocator: Add a postprocessing script for page-allocator-related ftrace events
> c33b3596bc38: tracing, page-allocator: Add trace event for page traffic related to the buddy lists
> 0d524fb734bc: tracing, mm: Add trace events for anti-fragmentation falling back to other migratetypes
> b9a28177eedf: tracing, page-allocator: Add trace events for page allocation and page freeing
> 807243eb20b2: Merge branch 'perfcounters/urgent' into tracing/mm
> 08b6cb88eeb5: perf_counter tools: Provide default bfd_demangle() function in case it's not around
> eb4671011887: tracing/mm: rename 'trigger' file to 'dump_range'
> 1487a7a1ff99: tracing/mm: fix mapcount trace record field
> dcac8cdac1d4: tracing/mm: add page frame snapshot trace
> 
> That's just a demo in essence - showing what things could be done in this area.
> 
> You can pick those commits up via running:
> 
>   http://people.redhat.com/mingo/tip.git/README

Ingo, sorry for remain silence on this topic. I'm actually updating the code.
Below is the most up-to-date version on top of 2.6.37.  I can post a patchset
for review after finished with the writeback patches.

Thanks,
Fengguang
---

Usage:

root@bay /home/wfg# echo / > /debug/tracing/objects/mm/pages/dump-fs
root@bay /home/wfg# cat /debug/tracing/trace

# The output has intermixed lines for inode and page
#         ino         size       cached      age(ms) dirty type first-opened-by file-name
      1507329         4096         8192       309042 ____  DIR          swapper /
#      index    len  page-flags count mapcount
           0      2 ____RU_____    1    0
      1786836        12288        40960       309026 ____  DIR          swapper /sbin
           0     10 ___ARU_____    1    0
      1786946        37312        40960       309024 ____  REG          swapper /sbin/init
           0      6 M__ARU_____    2    1
           6      1 M__A_U_____    2    1
           7      1 M__ARU_____    2    1
           8      2 _____U_____    1    0
      1507464            4         4096       309022 ____  LNK          swapper /lib64
           0      1 ___ARU_____    1    0
      1590173        12288            0       309021 ____  DIR          swapper /lib
      4563326           12         4096       309020 ____  LNK          swapper /lib/ld-linux-x86-64.so.2
           0      1 ___ARU_____    1    0
      4563295       128744       131072       309019 ____  REG          swapper /lib/ld-2.11.2.so
           0      1 M__ARU_____   21   20
           1      3 M__ARU_____   17   16
           4      4 M__ARU_____   20   19
           8      2 M__ARU_____   27   26
          10      3 M__ARU_____   20   19
          13      1 M__ARU_____   27   26
          14      1 M__ARU_____   26   25
          15      1 M__ARU_____   20   19
          16      1 M__ARU_____   18   17
          17      1 M__ARU_____    9    8
          18      1 M__A_U_____    4    3
          19      1 M__ARU_____   27   26
          20      1 M__ARU_____   17   16
          21      1 M__ARU_____   20   19
          22      1 M__ARU_____   27   26
          23      1 M__ARU_____   20   19
          24      1 M__ARU_____   26   25
          25      1 _____U_____    1    0
          26      1 M__A_U_____    4    3
          27      1 M__ARU_____   20   19
          28      4 _____U_____    1    0
      1525477        12288            0       309011 ____  DIR             init /etc
      1526463        64634        65536       309009 ____  REG             init /etc/ld.so.cache
           0      1 ___ARU_____    1    0
           1      1 _____U_____    1    0

--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ mmotm/include/trace/events/mm.h	2010-12-26 20:59:48.000000000 +0800
@@ -0,0 +1,164 @@
+#if !defined(_TRACE_MM_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_MM_H
+
+#include <linux/tracepoint.h>
+#include <linux/page-flags.h>
+#include <linux/memcontrol.h>
+#include <linux/pagemap.h>
+#include <linux/mm.h>
+#include <linux/kernel-page-flags.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM mm
+
+extern struct trace_print_flags pageflag_names[];
+
+/**
+ * dump_page_frame - called by the trace page dump trigger
+ * @pfn: page frame number
+ * @page: pointer to the page frame
+ *
+ * This is a helper trace point into the dumping of the page frames.
+ * It will record various infromation about a page frame.
+ */
+TRACE_EVENT(dump_page_frame,
+
+	TP_PROTO(unsigned long pfn, struct page *page),
+
+	TP_ARGS(pfn, page),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	pfn		)
+		__field(	struct page *,	page		)
+		__field(	u64,		stable_flags	)
+		__field(	unsigned long,	flags		)
+		__field(	unsigned int,	count		)
+		__field(	unsigned int,	mapcount	)
+		__field(	unsigned long,	private		)
+		__field(	unsigned long,	mapping		)
+		__field(	unsigned long,	index		)
+	),
+
+	TP_fast_assign(
+		__entry->pfn		= pfn;
+		__entry->page		= page;
+		__entry->stable_flags	= stable_page_flags(page);
+		__entry->flags		= page->flags;
+		__entry->count		= atomic_read(&page->_count);
+		__entry->mapcount	= page_mapcount(page);
+		__entry->private	= page->private;
+		__entry->mapping	= (unsigned long)page->mapping;
+		__entry->index		= page->index;
+	),
+
+	TP_printk("%12lx %16p %8x %8x %16lx %16lx %16lx %s",
+		  __entry->pfn,
+		  __entry->page,
+		  __entry->count,
+		  __entry->mapcount,
+		  __entry->private,
+		  __entry->mapping,
+		  __entry->index,
+		  ftrace_print_flags_seq(p, "|",
+					 __entry->flags & PAGE_FLAGS_MASK,
+					 pageflag_names)
+	)
+);
+
+TRACE_EVENT(dump_page_cache,
+
+	TP_PROTO(struct page *page, unsigned long len),
+
+	TP_ARGS(page, len),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	index		)
+		__field(	unsigned long,	len		)
+		__field(	u64,		flags		)
+		__field(	unsigned int,	count		)
+		__field(	unsigned int,	mapcount	)
+	),
+
+	TP_fast_assign(
+		__entry->index		= page->index;
+		__entry->len		= len;
+		__entry->flags		= stable_page_flags(page);
+		__entry->count		= atomic_read(&page->_count);
+		__entry->mapcount	= page_mapcount(page);
+	),
+
+	TP_printk("%12lu %6lu %c%c%c%c%c%c%c%c%c%c%c %4u %4u",
+		  __entry->index,
+		  __entry->len,
+		  __entry->flags & (1ULL << KPF_MMAP)		? 'M' : '_',
+		  __entry->flags & (1ULL << KPF_MLOCKED)	? 'm' : '_',
+		  __entry->flags & (1ULL << KPF_UNEVICTABLE)	? 'u' : '_',
+		  __entry->flags & (1ULL << KPF_ACTIVE)		? 'A' : '_',
+		  __entry->flags & (1ULL << KPF_REFERENCED)	? 'R' : '_',
+		  __entry->flags & (1ULL << KPF_UPTODATE)	? 'U' : '_',
+		  __entry->flags & (1ULL << KPF_DIRTY)		? 'D' : '_',
+		  __entry->flags & (1ULL << KPF_WRITEBACK)	? 'W' : '_',
+		  __entry->flags & (1ULL << KPF_RECLAIM)	? 'I' : '_',
+		  __entry->flags & (1ULL << KPF_MAPPEDTODISK)	? 'd' : '_',
+		  __entry->flags & (1ULL << KPF_PRIVATE)	? 'P' : '_',
+		  __entry->count,
+		  __entry->mapcount)
+);
+
+
+#define show_inode_type(val)	__print_symbolic(val, 	   \
+				{ S_IFREG,	"REG"	}, \
+				{ S_IFDIR,	"DIR"	}, \
+				{ S_IFLNK,	"LNK"	}, \
+				{ S_IFBLK,	"BLK"	}, \
+				{ S_IFCHR,	"CHR"	}, \
+				{ S_IFIFO,	"FIFO"	}, \
+				{ S_IFSOCK,	"SOCK"	})
+
+TRACE_EVENT(dump_inode_cache,
+
+	TP_PROTO(struct inode *inode, char *name, int len),
+
+	TP_ARGS(inode, name, len),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	ino		)
+		__field(	loff_t,		size		) /* bytes */
+		__field(	loff_t,		cached		) /* bytes */
+		__field(	unsigned long,	age		) /*    ms */
+		__field(	unsigned long,	state		)
+		__field(	umode_t,	mode		)
+		__array(	char,		comm, TASK_COMM_LEN)
+		__dynamic_array(char,		file,	len	)
+	),
+
+	TP_fast_assign(
+		__entry->ino	= inode->i_ino;
+		__entry->size	= i_size_read(inode);
+		__entry->cached	= inode->i_mapping->nrpages;
+		__entry->cached	<<= PAGE_CACHE_SHIFT;
+		__entry->age	= (jiffies - inode->dirtied_when) * 1000 / HZ;
+		__entry->state	= inode->i_state;
+		__entry->mode	= inode->i_mode;
+		memcpy(__entry->comm, inode->i_comm, TASK_COMM_LEN);
+		memcpy(__get_str(file), name, len);
+	),
+
+	TP_printk("%12lu %12llu %12llu %12lu %c%c%c%c %4s %16s %s",
+		  __entry->ino,
+		  __entry->size,
+		  __entry->cached,
+		  __entry->age,
+		  __entry->state & I_DIRTY_PAGES	? 'D' : '_',
+		  __entry->state & I_DIRTY_DATASYNC	? 'd' : '_',
+		  __entry->state & I_DIRTY_SYNC		? 'm' : '_',
+		  __entry->state & I_SYNC		? 'S' : '_',
+		  show_inode_type(__entry->mode & S_IFMT),
+		  __entry->comm,
+		  __get_str(file))
+);
+
+#endif /*  _TRACE_MM_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>
--- mmotm.orig/kernel/trace/Makefile	2010-12-26 20:58:46.000000000 +0800
+++ mmotm/kernel/trace/Makefile	2010-12-26 20:59:41.000000000 +0800
@@ -26,6 +26,7 @@ obj-$(CONFIG_RING_BUFFER) += ring_buffer
 obj-$(CONFIG_RING_BUFFER_BENCHMARK) += ring_buffer_benchmark.o
 
 obj-$(CONFIG_TRACING) += trace.o
+obj-$(CONFIG_TRACING) += trace_objects.o
 obj-$(CONFIG_TRACING) += trace_output.o
 obj-$(CONFIG_TRACING) += trace_stat.o
 obj-$(CONFIG_TRACING) += trace_printk.o
@@ -53,6 +54,7 @@ endif
 obj-$(CONFIG_EVENT_TRACING) += trace_events_filter.o
 obj-$(CONFIG_KPROBE_EVENT) += trace_kprobe.o
 obj-$(CONFIG_EVENT_TRACING) += power-traces.o
+obj-$(CONFIG_EVENT_TRACING) += trace_mm.o
 ifeq ($(CONFIG_TRACING),y)
 obj-$(CONFIG_KGDB_KDB) += trace_kdb.o
 endif
--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ mmotm/kernel/trace/trace_mm.c	2010-12-26 20:59:41.000000000 +0800
@@ -0,0 +1,367 @@
+/*
+ * Trace mm pages
+ *
+ * Copyright (C) 2009 Red Hat Inc, Steven Rostedt <srostedt@redhat.com>
+ *
+ * Code based on Matt Mackall's /proc/[kpagecount|kpageflags] code.
+ */
+#include <linux/module.h>
+#include <linux/bootmem.h>
+#include <linux/debugfs.h>
+#include <linux/uaccess.h>
+#include <linux/ctype.h>
+#include <linux/pagevec.h>
+#include <linux/writeback.h>
+#include <linux/file.h>
+#include <linux/slab.h>
+
+#include "trace_output.h"
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/mm.h>
+
+void trace_mm_page_frames(unsigned long start, unsigned long end,
+			  void (*trace)(unsigned long pfn, struct page *page))
+{
+	unsigned long pfn = start;
+	struct page *page;
+
+	if (start > max_pfn - 1)
+		return;
+
+	if (end > max_pfn)
+		end = max_pfn;
+
+	while (pfn < end) {
+		page = NULL;
+		if (pfn_valid(pfn))
+			page = pfn_to_page(pfn);
+		pfn++;
+		if (page)
+			trace(pfn, page);
+	}
+}
+
+static void trace_mm_page_frame(unsigned long pfn, struct page *page)
+{
+	trace_dump_page_frame(pfn, page);
+}
+
+static ssize_t
+trace_mm_pfn_range_read(struct file *filp, char __user *ubuf, size_t cnt,
+			loff_t *ppos)
+{
+	return simple_read_from_buffer(ubuf, cnt, ppos, "0\n", 2);
+}
+
+
+/*
+ * recognized formats:
+ * 		"M N"	start=M, end=N
+ * 		"M"	start=M, end=M+1
+ * 		"M +N"	start=M, end=M+N-1
+ */
+static ssize_t
+trace_mm_pfn_range_write(struct file *filp, const char __user *ubuf, size_t cnt,
+			 loff_t *ppos)
+{
+	unsigned long start;
+	unsigned long end = 0;
+	char buf[64];
+	char *ptr;
+
+	if (cnt >= sizeof(buf))
+		return -EINVAL;
+
+	if (copy_from_user(&buf, ubuf, cnt))
+		return -EFAULT;
+
+	if (tracing_update_buffers() < 0)
+		return -ENOMEM;
+
+	if (trace_set_clr_event("mm", "dump_page_frame", 1))
+		return -EINVAL;
+
+	buf[cnt] = 0;
+
+	start = simple_strtoul(buf, &ptr, 0);
+
+	for (; *ptr; ptr++) {
+		if (isdigit(*ptr)) {
+			if (*(ptr - 1) == '+')
+				end = start;
+			end += simple_strtoul(ptr, NULL, 0);
+			break;
+		}
+	}
+	if (!*ptr)
+		end = start + 1;
+
+	trace_mm_page_frames(start, end, trace_mm_page_frame);
+
+	return cnt;
+}
+
+static const struct file_operations trace_mm_fops = {
+	.open		= tracing_open_generic,
+	.read		= trace_mm_pfn_range_read,
+	.write		= trace_mm_pfn_range_write,
+};
+
+static struct dentry *trace_objects_mm_dir(void)
+{
+	static struct dentry *d_mm;
+	struct dentry *d_objects;
+
+	if (d_mm)
+		return d_mm;
+
+	d_objects = trace_objects_dir();
+	if (!d_objects)
+		return NULL;
+
+	d_mm = debugfs_create_dir("mm", d_objects);
+	if (!d_mm)
+		pr_warning("Could not create 'objects/mm' directory\n");
+
+	return d_mm;
+}
+
+static unsigned long page_flags(struct page *page)
+{
+	return page->flags & ((1 << NR_PAGEFLAGS) - 1);
+}
+
+static int pages_similar(struct page *page0, struct page *page)
+{
+	if (page_flags(page0) != page_flags(page))
+		return 0;
+
+	if (page_count(page0) != page_count(page))
+		return 0;
+
+	if (page_mapcount(page0) != page_mapcount(page))
+		return 0;
+
+	return 1;
+}
+
+static void dump_pagecache(struct address_space *mapping)
+{
+	unsigned long nr_pages;
+	struct page *pages[PAGEVEC_SIZE];
+	struct page *uninitialized_var(page0);
+	struct page *page;
+	unsigned long start = 0;
+	unsigned long len = 0;
+	int i;
+
+	for (;;) {
+		rcu_read_lock();
+		nr_pages = radix_tree_gang_lookup(&mapping->page_tree,
+				(void **)pages, start + len, PAGEVEC_SIZE);
+		rcu_read_unlock();
+
+		if (nr_pages == 0) {
+			if (len)
+				trace_dump_page_cache(page0, len);
+			return;
+		}
+
+		for (i = 0; i < nr_pages; i++) {
+			page = pages[i];
+
+			if (len &&
+			    page->index == start + len &&
+			    pages_similar(page0, page))
+				len++;
+			else {
+				if (len)
+					trace_dump_page_cache(page0, len);
+				page0 = page;
+				start = page->index;
+				len = 1;
+			}
+		}
+		cond_resched();
+	}
+}
+
+static void dump_inode_cache(struct inode *inode,
+			     char *name_buf,
+			     struct vfsmount *mnt)
+{
+	struct path path = {
+		.mnt = mnt,
+		.dentry = d_find_alias(inode)
+	};
+	char *name;
+	int len;
+
+	if (!mnt) {
+		trace_dump_inode_cache(inode, name_buf, strlen(name_buf));
+		return;
+	}
+
+	if (!path.dentry) {
+		trace_dump_inode_cache(inode, "", 1);
+		return;
+	}
+
+	name = d_path(&path, name_buf, PAGE_SIZE);
+	if (IS_ERR(name)) {
+		name = "";
+		len = 1;
+	} else
+		len = PAGE_SIZE + name_buf - name;
+
+	trace_dump_inode_cache(inode, name, len);
+
+	if (path.dentry)
+		dput(path.dentry);
+}
+
+static void dump_fs_pagecache(struct super_block *sb, struct vfsmount *mnt)
+{
+	struct inode *inode;
+	struct inode *prev_inode = NULL;
+	char *name_buf;
+
+	name_buf = (char *)__get_free_page(GFP_TEMPORARY);
+	if (!name_buf)
+		return;
+
+	down_read(&sb->s_umount);
+	if (!sb->s_root)
+		goto out;
+
+	spin_lock(&inode_lock);
+	list_for_each_entry_reverse(inode, &sb->s_inodes, i_sb_list) {
+		if (inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE|I_NEW))
+			continue;
+		__iget(inode);
+		spin_unlock(&inode_lock);
+		dump_inode_cache(inode, name_buf, mnt);
+		if (inode->i_mapping->nrpages)
+			dump_pagecache(inode->i_mapping);
+		iput(prev_inode);
+		prev_inode = inode;
+		cond_resched();
+		spin_lock(&inode_lock);
+	}
+	spin_unlock(&inode_lock);
+	iput(prev_inode);
+out:
+	up_read(&sb->s_umount);
+	free_page((unsigned long)name_buf);
+}
+
+static ssize_t
+trace_pagecache_write(struct file *filp, const char __user *ubuf, size_t count,
+		      loff_t *ppos)
+{
+	struct file *file = NULL;
+	char *name;
+	int err = 0;
+
+	if (count <= 1)
+		return -EINVAL;
+	if (count >= PAGE_SIZE)
+		return -ENAMETOOLONG;
+
+	name = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!name)
+		return -ENOMEM;
+
+	if (copy_from_user(name, ubuf, count)) {
+		err = -EFAULT;
+		goto out;
+	}
+
+	/* strip the newline added by `echo` */
+	if (name[count-1] == '\n')
+		name[count-1] = '\0';
+	else
+		name[count] = '\0';
+
+	file = filp_open(name, O_RDONLY|O_LARGEFILE, 0);
+	if (IS_ERR(file)) {
+		err = PTR_ERR(file);
+		file = NULL;
+		goto out;
+	}
+
+	if (tracing_update_buffers() < 0) {
+		err = -ENOMEM;
+		goto out;
+	}
+	if (trace_set_clr_event("mm", "dump_page_cache", 1)) {
+		err = -EINVAL;
+		goto out;
+	}
+	if (trace_set_clr_event("mm", "dump_inode_cache", 1)) {
+		err = -EINVAL;
+		goto out;
+	}
+
+	if (filp->f_path.dentry->d_inode->i_private) {
+		dump_fs_pagecache(file->f_path.dentry->d_sb, file->f_path.mnt);
+	} else {
+		dump_inode_cache(file->f_mapping->host, name, NULL);
+		dump_pagecache(file->f_mapping);
+	}
+
+out:
+	if (file)
+		fput(file);
+	kfree(name);
+
+	return err ? err : count;
+}
+
+static const struct file_operations trace_pagecache_fops = {
+	.open		= tracing_open_generic,
+	.read		= trace_mm_pfn_range_read,
+	.write		= trace_pagecache_write,
+};
+
+static struct dentry *trace_objects_mm_pages_dir(void)
+{
+	static struct dentry *d_pages;
+	struct dentry *d_mm;
+
+	if (d_pages)
+		return d_pages;
+
+	d_mm = trace_objects_mm_dir();
+	if (!d_mm)
+		return NULL;
+
+	d_pages = debugfs_create_dir("pages", d_mm);
+	if (!d_pages)
+		pr_warning("Could not create debugfs "
+			   "'objects/mm/pages' directory\n");
+
+	return d_pages;
+}
+
+static __init int trace_objects_mm_init(void)
+{
+	struct dentry *d_pages;
+
+	d_pages = trace_objects_mm_pages_dir();
+	if (!d_pages)
+		return 0;
+
+	trace_create_file("dump-pfn", 0600, d_pages, NULL,
+			  &trace_mm_fops);
+
+	trace_create_file("dump-file", 0600, d_pages, NULL,
+			  &trace_pagecache_fops);
+
+	trace_create_file("dump-fs", 0600, d_pages, (void *)1,
+			  &trace_pagecache_fops);
+
+	return 0;
+}
+fs_initcall(trace_objects_mm_init);
--- mmotm.orig/kernel/trace/trace.h	2010-12-26 20:58:46.000000000 +0800
+++ mmotm/kernel/trace/trace.h	2010-12-26 20:59:41.000000000 +0800
@@ -295,6 +295,7 @@ struct dentry *trace_create_file(const c
 				 const struct file_operations *fops);
 
 struct dentry *tracing_init_dentry(void);
+struct dentry *trace_objects_dir(void);
 
 struct ring_buffer_event;
 
--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ mmotm/kernel/trace/trace_objects.c	2010-12-26 20:59:41.000000000 +0800
@@ -0,0 +1,26 @@
+#include <linux/debugfs.h>
+
+#include "trace.h"
+#include "trace_output.h"
+
+struct dentry *trace_objects_dir(void)
+{
+	static struct dentry *d_objects;
+	struct dentry *d_tracer;
+
+	if (d_objects)
+		return d_objects;
+
+	d_tracer = tracing_init_dentry();
+	if (!d_tracer)
+		return NULL;
+
+	d_objects = debugfs_create_dir("objects", d_tracer);
+	if (!d_objects)
+		pr_warning("Could not create debugfs "
+			   "'objects' directory\n");
+
+	return d_objects;
+}
+
+
--- mmotm.orig/mm/page_alloc.c	2010-12-26 20:58:46.000000000 +0800
+++ mmotm/mm/page_alloc.c	2010-12-26 20:59:41.000000000 +0800
@@ -5493,7 +5493,7 @@ bool is_free_buddy_page(struct page *pag
 }
 #endif
 
-static struct trace_print_flags pageflag_names[] = {
+struct trace_print_flags pageflag_names[] = {
 	{1UL << PG_locked,		"locked"	},
 	{1UL << PG_error,		"error"		},
 	{1UL << PG_referenced,		"referenced"	},
@@ -5541,7 +5541,7 @@ static void dump_page_flags(unsigned lon
 	printk(KERN_ALERT "page flags: %#lx(", flags);
 
 	/* remove zone id */
-	flags &= (1UL << NR_PAGEFLAGS) - 1;
+	flags &= PAGE_FLAGS_MASK;
 
 	for (i = 0; pageflag_names[i].name && flags; i++) {
 
--- mmotm.orig/include/linux/page-flags.h	2010-12-26 20:58:46.000000000 +0800
+++ mmotm/include/linux/page-flags.h	2010-12-26 20:59:41.000000000 +0800
@@ -414,6 +414,7 @@ static inline void __ClearPageTail(struc
  * there has been a kernel bug or struct page corruption.
  */
 #define PAGE_FLAGS_CHECK_AT_PREP	((1 << NR_PAGEFLAGS) - 1)
+#define PAGE_FLAGS_MASK			((1 << NR_PAGEFLAGS) - 1)
 
 #define PAGE_FLAGS_PRIVATE				\
 	(1 << PG_private | 1 << PG_private_2)
--- mmotm.orig/fs/inode.c	2010-12-26 20:58:45.000000000 +0800
+++ mmotm/fs/inode.c	2010-12-26 21:00:09.000000000 +0800
@@ -182,7 +182,13 @@ int inode_init_always(struct super_block
 	inode->i_bdev = NULL;
 	inode->i_cdev = NULL;
 	inode->i_rdev = 0;
-	inode->dirtied_when = 0;
+
+	/*
+	 * This records inode load time. It will be invalidated once inode is
+	 * dirtied, or jiffies wraps around. Despite the pitfalls it still
+	 * provides useful information for some use cases like fastboot.
+	 */
+	inode->dirtied_when = jiffies;
 
 	if (security_inode_alloc(inode))
 		goto out;
@@ -226,6 +232,9 @@ int inode_init_always(struct super_block
 
 	percpu_counter_inc(&nr_inodes);
 
+	BUILD_BUG_ON(sizeof(inode->i_comm) != TASK_COMM_LEN);
+	memcpy(inode->i_comm, current->comm, TASK_COMM_LEN);
+
 	return 0;
 out:
 	return -ENOMEM;
--- mmotm.orig/include/linux/fs.h	2010-12-26 20:59:50.000000000 +0800
+++ mmotm/include/linux/fs.h	2010-12-26 21:00:09.000000000 +0800
@@ -800,6 +800,8 @@ struct inode {
 	struct posix_acl	*i_default_acl;
 #endif
 	void			*i_private; /* fs or device private pointer */
+
+	char			i_comm[16]; /* first opened by */
 };
 
 static inline int inode_unhashed(struct inode *inode)

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[KOSAKI Motohiro]

> 
> * Liu Yuan <namei.unix@gmail.com> wrote:
> 
> > +		if (likely(!retry_find) && page && PageUptodate(page))
> > +			page_cache_acct_hit(inode->i_sb, READ);
> > +		else
> > +			page_cache_acct_missed(inode->i_sb, READ);
> 
> Sigh.
> 
> This would make such a nice tracepoint or sw perf event. It could be collected in a 
> 'count' form, equivalent to the stats you are aiming for here, or it could even be 
> traced, if someone is interested in such details.
> 
> It could be mixed with other events, enriching multiple apps at once.

Totally agreed.


> But, instead of trying to improve those aspects of our existing instrumentation 
> frameworks, mm/* is gradually growing its own special instrumentation hacks, missing 
> the big picture and fragmenting the instrumentation space some more.
> 
> That trend is somewhat sad.

So, I think thing is, the stat is how much people and how frequently used.
If it is really really really common, /proc/meminfo or similar special place
is good idea. Another example, If the stat can help our MM debugging to
handle LKML bug report, it is worth to have special care. But other almost
else case are better to use generic framework.

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Tao Ma]

On 03/02/2011 04:45 PM, Ingo Molnar wrote:
> * Liu Yuan<namei.unix@gmail.com>  wrote:
>
>    
>> +		if (likely(!retry_find)&&  page&&  PageUptodate(page))
>> +			page_cache_acct_hit(inode->i_sb, READ);
>> +		else
>> +			page_cache_acct_missed(inode->i_sb, READ);
>>      
> Sigh.
>
> This would make such a nice tracepoint or sw perf event. It could be collected in a
> 'count' form, equivalent to the stats you are aiming for here, or it could even be
> traced, if someone is interested in such details.
>
> It could be mixed with other events, enriching multiple apps at once.
>
> But, instead of trying to improve those aspects of our existing instrumentation
> frameworks, mm/* is gradually growing its own special instrumentation hacks, missing
> the big picture and fragmenting the instrumentation space some more.
>    
Thanks for the quick response. Actually our team(including Liu) here are 
planing to add some
debug info to the mm parts for analyzing the application behavior and 
hope to find some way
to improve our application's performance.
We have searched the trace points in mm, but it seems to us that the 
trace points isn't quite welcomed
there. Only vmscan and writeback have some limited trace points added. 
That's the reason we first
tried to add some debug info like this patch. You does shed some light 
on our direction. Thanks.

btw, what part do you think is needed to add some trace point?  We 
volunteer to add more if you like.

Regards,
Tao

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Ingo Molnar]

* Tao Ma <tm@tao.ma> wrote:

> On 03/02/2011 04:45 PM, Ingo Molnar wrote:
> >* Liu Yuan<namei.unix@gmail.com>  wrote:
> >
> >>+		if (likely(!retry_find)&&  page&&  PageUptodate(page))
> >>+			page_cache_acct_hit(inode->i_sb, READ);
> >>+		else
> >>+			page_cache_acct_missed(inode->i_sb, READ);
> >Sigh.
> >
> >This would make such a nice tracepoint or sw perf event. It could be collected in a
> >'count' form, equivalent to the stats you are aiming for here, or it could even be
> >traced, if someone is interested in such details.
> >
> >It could be mixed with other events, enriching multiple apps at once.
> >
> >But, instead of trying to improve those aspects of our existing instrumentation
> >frameworks, mm/* is gradually growing its own special instrumentation hacks, missing
> >the big picture and fragmenting the instrumentation space some more.

> Thanks for the quick response. Actually our team(including Liu) here are planing 
> to add some debug info to the mm parts for analyzing the application behavior and 
> hope to find some way to improve our application's performance. We have searched 
> the trace points in mm, but it seems to us that the trace points isn't quite 
> welcomed there. Only vmscan and writeback have some limited trace points added. 
> That's the reason we first tried to add some debug info like this patch. You does 
> shed some light on our direction. Thanks.

Yes, it's very much a 'critical mass' phenomenon: the moment there's enough 
tracepoints, above some magic limit, things happen quickly and everyone finds the 
stuff obviously useful.

Before that limit it's all pretty painful.

> btw, what part do you think is needed to add some trace point?  We
> volunteer to add more if you like.

Whatever part you find useful in your daily development work!

Tracepoints are pretty flexible. The bit that is missing and which is very important 
for the MM is the collapse into 'summaries' and the avoidance of tracing overhead 
when only a summary is wanted. Please see Wu Fengguang's reply in this thread about 
the 'dump state' facility he and Steve added to recover large statistics.

I suspect the hit/miss histogram you are building in this patch could be recovered 
via that facility initially?

The next step would generalize that approach - it is non-trivial but powerful :-)

The idea is to allow non-trivial histograms and summaries to be built out of simple 
events, via the filter engine.

It would require an extension of tracing to really allow a filter expression to be 
defined over existing events, which would allow the maintenance of a persistent 
'sum' variable - probably within the perf ring-buffer. We already have filter 
support, that would have to be extended with a notion of 'persistent variables'.

So right now, if you define a tracepoint in that spot, we already support such 
filter expressions:

     'bdev == sda1 && page_state == PageUptodate'

You can inject such filter expressions into /debug/tracing/events/*/*/filter today, 
and you can use filters in perf record --filter '...' as well.

To implement 'fast statistics', the filter engine would have to be extended to 
support (simple) statements like:

	if (bdev == sda1 && page_state == PageUptodate)'
		var0++;

And:

	if (bdev == sda1 && page_state != PageUptodate)'
		var1++;

Only a very minimal type of C syntax would be supported - not a full C parser.

That way the 'var0' portion of the perf ring-buffer (which would not be part of the 
regular, overwritten ring-buffer) would act as a 'hits' variable that you could 
recover. The 'var1' portion would be the 'misses' counter.

Individual trace events would only twiddle var0 and var1 - they would not inject a 
full-blown event into the ring-buffer, so statistics would be very fast.

This method is very extensible and could be used for far more things than just MM 
statistics. In theory all of /proc statistics collection could be replaced and made 
optional that way, just by adding the right events to the right spots in the kernel.  
That is obviously a very long-term project.

Thanks,

	Ingo

Re: [RFC PATCH 4/5] mm: Add hit/miss accounting for Page Cache

[Tao Ma]

On 03/03/2011 05:34 PM, Ingo Molnar wrote:
> * Tao Ma<tm@tao.ma>  wrote:
>
>> On 03/02/2011 04:45 PM, Ingo Molnar wrote:
>>> * Liu Yuan<namei.unix@gmail.com>   wrote:
>>>
>>>> +		if (likely(!retry_find)&&   page&&   PageUptodate(page))
>>>> +			page_cache_acct_hit(inode->i_sb, READ);
>>>> +		else
>>>> +			page_cache_acct_missed(inode->i_sb, READ);
>>> Sigh.
>>>
>>> This would make such a nice tracepoint or sw perf event. It could be collected in a
>>> 'count' form, equivalent to the stats you are aiming for here, or it could even be
>>> traced, if someone is interested in such details.
>>>
>>> It could be mixed with other events, enriching multiple apps at once.
>>>
>>> But, instead of trying to improve those aspects of our existing instrumentation
>>> frameworks, mm/* is gradually growing its own special instrumentation hacks, missing
>>> the big picture and fragmenting the instrumentation space some more.
>> Thanks for the quick response. Actually our team(including Liu) here are planing
>> to add some debug info to the mm parts for analyzing the application behavior and
>> hope to find some way to improve our application's performance. We have searched
>> the trace points in mm, but it seems to us that the trace points isn't quite
>> welcomed there. Only vmscan and writeback have some limited trace points added.
>> That's the reason we first tried to add some debug info like this patch. You does
>> shed some light on our direction. Thanks.
> Yes, it's very much a 'critical mass' phenomenon: the moment there's enough
> tracepoints, above some magic limit, things happen quickly and everyone finds the
> stuff obviously useful.
>
> Before that limit it's all pretty painful.
yeah.
>> btw, what part do you think is needed to add some trace point?  We
>> volunteer to add more if you like.
> Whatever part you find useful in your daily development work!
>
> Tracepoints are pretty flexible. The bit that is missing and which is very important
> for the MM is the collapse into 'summaries' and the avoidance of tracing overhead
> when only a summary is wanted. Please see Wu Fengguang's reply in this thread about
> the 'dump state' facility he and Steve added to recover large statistics.
We are looking into it now. Thanks for the hint.
> I suspect the hit/miss histogram you are building in this patch could be recovered
> via that facility initially?
>
> The next step would generalize that approach - it is non-trivial but powerful :-)
>
> The idea is to allow non-trivial histograms and summaries to be built out of simple
> events, via the filter engine.
>
> It would require an extension of tracing to really allow a filter expression to be
> defined over existing events, which would allow the maintenance of a persistent
> 'sum' variable - probably within the perf ring-buffer. We already have filter
> support, that would have to be extended with a notion of 'persistent variables'.
>
> So right now, if you define a tracepoint in that spot, we already support such
> filter expressions:
>
>       'bdev == sda1&&  page_state == PageUptodate'
>
> You can inject such filter expressions into /debug/tracing/events/*/*/filter today,
> and you can use filters in perf record --filter '...' as well.
>
> To implement 'fast statistics', the filter engine would have to be extended to
> support (simple) statements like:
>
> 	if (bdev == sda1&&  page_state == PageUptodate)'
> 		var0++;
>
> And:
>
> 	if (bdev == sda1&&  page_state != PageUptodate)'
> 		var1++;
>
> Only a very minimal type of C syntax would be supported - not a full C parser.
>
> That way the 'var0' portion of the perf ring-buffer (which would not be part of the
> regular, overwritten ring-buffer) would act as a 'hits' variable that you could
> recover. The 'var1' portion would be the 'misses' counter.
>
> Individual trace events would only twiddle var0 and var1 - they would not inject a
> full-blown event into the ring-buffer, so statistics would be very fast.
>
> This method is very extensible and could be used for far more things than just MM
> statistics. In theory all of /proc statistics collection could be replaced and made
> optional that way, just by adding the right events to the right spots in the kernel.
> That is obviously a very long-term project.
It looks really fantastic for us. OK, we will try to figure out when and 
how we can work on this issue. Great thanks.

Regards,
Tao

[PATCH] fixed a macro coding style issue

[Baodong Chen]

Fixed a coding style issue in driver/input/input.c

Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
---
 drivers/input/input.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/input/input.c b/drivers/input/input.c
index 8921c61..c96e983 100644
--- a/drivers/input/input.c
+++ b/drivers/input/input.c
@@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
 EXPORT_SYMBOL(input_set_keycode);
 
 #define MATCH_BIT(bit, max) \
+	do { \
 		for (i = 0; i < BITS_TO_LONGS(max); i++) \
 			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
 				break; \
-		if (i != BITS_TO_LONGS(max)) \
-			continue;
+			if (i != BITS_TO_LONGS(max)) \
+				continue; \
+	} while (0)
 
 static const struct input_device_id *input_match_device(struct input_handler *handler,
 							struct input_dev *dev)
-- 
1.7.0.4

RE: [PATCH] fixed a macro coding style issue

[Venu Byravarasu]

> -----Original Message-----
> From: linux-kernel-owner@vger.kernel.org [mailto:linux-kernel-
> owner@vger.kernel.org] On Behalf Of Baodong Chen
> Sent: Wednesday, July 25, 2012 10:51 AM
> To: dmitry.torokhov@gmail.com
> Cc: linux-input@vger.kernel.org; linux-kernel@vger.kernel.org; Baodong
> Chen
> Subject: [PATCH] fixed a macro coding style issue
> 
> Fixed a coding style issue in driver/input/input.c
> 
> Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
> ---
>  drivers/input/input.c |    6 ++++--
>  1 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/input/input.c b/drivers/input/input.c
> index 8921c61..c96e983 100644
> --- a/drivers/input/input.c
> +++ b/drivers/input/input.c
> @@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
>  EXPORT_SYMBOL(input_set_keycode);
> 
>  #define MATCH_BIT(bit, max) \
> +	do { \
>  		for (i = 0; i < BITS_TO_LONGS(max); i++) \
>  			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
>  				break; \
> -		if (i != BITS_TO_LONGS(max)) \
> -			continue;
> +			if (i != BITS_TO_LONGS(max)) \
> +				continue; \

What is the need for extra indentation here?

> +	} while (0)
> 
>  static const struct input_device_id *input_match_device(struct
> input_handler *handler,
>  							struct input_dev
> *dev)
> --
> 1.7.0.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH] fixed a macro coding style issue

[Dmitry Torokhov]

On Wed, Jul 25, 2012 at 01:20:56PM +0800, Baodong Chen wrote:
> Fixed a coding style issue in driver/input/input.c
> 
> Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
> ---
>  drivers/input/input.c |    6 ++++--
>  1 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/input/input.c b/drivers/input/input.c
> index 8921c61..c96e983 100644
> --- a/drivers/input/input.c
> +++ b/drivers/input/input.c
> @@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
>  EXPORT_SYMBOL(input_set_keycode);
>  
>  #define MATCH_BIT(bit, max) \
> +	do { \
>  		for (i = 0; i < BITS_TO_LONGS(max); i++) \
>  			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
>  				break; \
> -		if (i != BITS_TO_LONGS(max)) \
> -			continue;
> +			if (i != BITS_TO_LONGS(max)) \
> +				continue; \
> +	} while (0)

This changes semantics. While current implementation of MATCH_BIT might
not be great style, it is not supposed to be used outside of
input_match_device(), and the replacement is completely broken.

Thanks.

-- 
Dmitry

Re: [PATCH] fixed a macro coding style issue

[Baodong Chen]

this patch definitely changes semantics.
so please discard it, sorry about that!

On Wed, Jul 25, 2012 at 1:37 PM, Dmitry Torokhov
<dmitry.torokhov@gmail.com> wrote:
> On Wed, Jul 25, 2012 at 01:20:56PM +0800, Baodong Chen wrote:
>> Fixed a coding style issue in driver/input/input.c
>>
>> Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
>> ---
>>  drivers/input/input.c |    6 ++++--
>>  1 files changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/input/input.c b/drivers/input/input.c
>> index 8921c61..c96e983 100644
>> --- a/drivers/input/input.c
>> +++ b/drivers/input/input.c
>> @@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
>>  EXPORT_SYMBOL(input_set_keycode);
>>
>>  #define MATCH_BIT(bit, max) \
>> +     do { \
>>               for (i = 0; i < BITS_TO_LONGS(max); i++) \
>>                       if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
>>                               break; \
>> -             if (i != BITS_TO_LONGS(max)) \
>> -                     continue;
>> +                     if (i != BITS_TO_LONGS(max)) \
>> +                             continue; \
>> +     } while (0)
>
> This changes semantics. While current implementation of MATCH_BIT might
> not be great style, it is not supposed to be used outside of
> input_match_device(), and the replacement is completely broken.
>
> Thanks.
>
> --
> Dmitry

Re: [PATCH] fixed a macro coding style issue

[Al Viro]

On Tue, Jul 24, 2012 at 10:37:55PM -0700, Dmitry Torokhov wrote:
> On Wed, Jul 25, 2012 at 01:20:56PM +0800, Baodong Chen wrote:
> > Fixed a coding style issue in driver/input/input.c
> > 
> > Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
> > ---
> >  drivers/input/input.c |    6 ++++--
> >  1 files changed, 4 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/input/input.c b/drivers/input/input.c
> > index 8921c61..c96e983 100644
> > --- a/drivers/input/input.c
> > +++ b/drivers/input/input.c
> > @@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
> >  EXPORT_SYMBOL(input_set_keycode);
> >  
> >  #define MATCH_BIT(bit, max) \
> > +	do { \
> >  		for (i = 0; i < BITS_TO_LONGS(max); i++) \
> >  			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
> >  				break; \
> > -		if (i != BITS_TO_LONGS(max)) \
> > -			continue;
> > +			if (i != BITS_TO_LONGS(max)) \
> > +				continue; \
> > +	} while (0)
> 
> This changes semantics. While current implementation of MATCH_BIT might
> not be great style, it is not supposed to be used outside of
> input_match_device(), and the replacement is completely broken.

Replacement is certainly completely broken, but "might not be great style"
is quite an understatement.  Everything else aside, it's too obfuscated
and ugly for anything outside of IOCCC and too verbose for IOCCC.

Damnit, at least turn that into
static inline int is_subset(unsigned long *bitmap1, unsigned long *bitmap2, size_t bits)
{
	int i;
	for (i = BITS_TO_LONGS(bits); i; bitmap1++, bitmap2++, i--)
		if ((*bitmap1 & *bitmap2) != *bitmap1)
			return 0;
	return 1;
}
and just before your loop
/* ignore ones that don't have bits required by id */
#define MATCH_BIT(array, bits) if (!is_subset(id->array, dev->array, bits)) continue;
with #undef MATCH_BIT right after the loop, to make it damn clean that it's local.
And frankly, I'd consider expanding that variant and killing the macro off.
Macros that affect control flow are vile and actively confusing for reader.

Re: [PATCH] fixed a macro coding style issue

[Dmitry Torokhov]

On Wed, Jul 25, 2012 at 07:15:12AM +0100, Al Viro wrote:
> On Tue, Jul 24, 2012 at 10:37:55PM -0700, Dmitry Torokhov wrote:
> > On Wed, Jul 25, 2012 at 01:20:56PM +0800, Baodong Chen wrote:
> > > Fixed a coding style issue in driver/input/input.c
> > > 
> > > Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
> > > ---
> > >  drivers/input/input.c |    6 ++++--
> > >  1 files changed, 4 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/drivers/input/input.c b/drivers/input/input.c
> > > index 8921c61..c96e983 100644
> > > --- a/drivers/input/input.c
> > > +++ b/drivers/input/input.c
> > > @@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
> > >  EXPORT_SYMBOL(input_set_keycode);
> > >  
> > >  #define MATCH_BIT(bit, max) \
> > > +	do { \
> > >  		for (i = 0; i < BITS_TO_LONGS(max); i++) \
> > >  			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
> > >  				break; \
> > > -		if (i != BITS_TO_LONGS(max)) \
> > > -			continue;
> > > +			if (i != BITS_TO_LONGS(max)) \
> > > +				continue; \
> > > +	} while (0)
> > 
> > This changes semantics. While current implementation of MATCH_BIT might
> > not be great style, it is not supposed to be used outside of
> > input_match_device(), and the replacement is completely broken.
> 
> Replacement is certainly completely broken, but "might not be great style"
> is quite an understatement.  Everything else aside, it's too obfuscated
> and ugly for anything outside of IOCCC and too verbose for IOCCC.
> 
> Damnit, at least turn that into
> static inline int is_subset(unsigned long *bitmap1, unsigned long *bitmap2, size_t bits)
> {
> 	int i;
> 	for (i = BITS_TO_LONGS(bits); i; bitmap1++, bitmap2++, i--)
> 		if ((*bitmap1 & *bitmap2) != *bitmap1)
> 			return 0;
> 	return 1;
> }
> and just before your loop
> /* ignore ones that don't have bits required by id */
> #define MATCH_BIT(array, bits) if (!is_subset(id->array, dev->array, bits)) continue;

Hmm, we already have bitmap_subset(), I think we can simply use it.

> with #undef MATCH_BIT right after the loop, to make it damn clean that it's local.
> And frankly, I'd consider expanding that variant and killing the macro off.
> Macros that affect control flow are vile and actively confusing for reader.

I agree. In my defense it is quite old code...

-- 
Dmitry

Re: [PATCH] fixed a macro coding style issue

[Dmitry Torokhov]

On Tue, Jul 24, 2012 at 11:36:24PM -0700, Dmitry Torokhov wrote:
> On Wed, Jul 25, 2012 at 07:15:12AM +0100, Al Viro wrote:
> > On Tue, Jul 24, 2012 at 10:37:55PM -0700, Dmitry Torokhov wrote:
> > > On Wed, Jul 25, 2012 at 01:20:56PM +0800, Baodong Chen wrote:
> > > > Fixed a coding style issue in driver/input/input.c
> > > > 
> > > > Signed-off-by: Baodong Chen <chenbdchenbd@gmail.com>
> > > > ---
> > > >  drivers/input/input.c |    6 ++++--
> > > >  1 files changed, 4 insertions(+), 2 deletions(-)
> > > > 
> > > > diff --git a/drivers/input/input.c b/drivers/input/input.c
> > > > index 8921c61..c96e983 100644
> > > > --- a/drivers/input/input.c
> > > > +++ b/drivers/input/input.c
> > > > @@ -845,11 +845,13 @@ int input_set_keycode(struct input_dev *dev,
> > > >  EXPORT_SYMBOL(input_set_keycode);
> > > >  
> > > >  #define MATCH_BIT(bit, max) \
> > > > +	do { \
> > > >  		for (i = 0; i < BITS_TO_LONGS(max); i++) \
> > > >  			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
> > > >  				break; \
> > > > -		if (i != BITS_TO_LONGS(max)) \
> > > > -			continue;
> > > > +			if (i != BITS_TO_LONGS(max)) \
> > > > +				continue; \
> > > > +	} while (0)
> > > 
> > > This changes semantics. While current implementation of MATCH_BIT might
> > > not be great style, it is not supposed to be used outside of
> > > input_match_device(), and the replacement is completely broken.
> > 
> > Replacement is certainly completely broken, but "might not be great style"
> > is quite an understatement.  Everything else aside, it's too obfuscated
> > and ugly for anything outside of IOCCC and too verbose for IOCCC.
> > 
> > Damnit, at least turn that into
> > static inline int is_subset(unsigned long *bitmap1, unsigned long *bitmap2, size_t bits)
> > {
> > 	int i;
> > 	for (i = BITS_TO_LONGS(bits); i; bitmap1++, bitmap2++, i--)
> > 		if ((*bitmap1 & *bitmap2) != *bitmap1)
> > 			return 0;
> > 	return 1;
> > }
> > and just before your loop
> > /* ignore ones that don't have bits required by id */
> > #define MATCH_BIT(array, bits) if (!is_subset(id->array, dev->array, bits)) continue;
> 
> Hmm, we already have bitmap_subset(), I think we can simply use it.
> 
> > with #undef MATCH_BIT right after the loop, to make it damn clean that it's local.
> > And frankly, I'd consider expanding that variant and killing the macro off.
> > Macros that affect control flow are vile and actively confusing for reader.
> 
> I agree. In my defense it is quite old code...

So how about the patch below?

-- 
Dmitry

Input: get rid of MATCH_BIT() macro

From: Dmitry Torokhov <dmitry.torokhov@gmail.com>

MATCH_BIT() is ugly and stupid, we have much nicer bitmap_subset() which
does the same and does not hide control flow.

Reported-by: Baodong Chen <chenbdchenbd@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---

 drivers/input/input.c |   43 ++++++++++++++++++++++++++-----------------
 1 file changed, 26 insertions(+), 17 deletions(-)


diff --git a/drivers/input/input.c b/drivers/input/input.c
index 8921c61..768e46b 100644
--- a/drivers/input/input.c
+++ b/drivers/input/input.c
@@ -844,18 +844,10 @@ int input_set_keycode(struct input_dev *dev,
 }
 EXPORT_SYMBOL(input_set_keycode);
 
-#define MATCH_BIT(bit, max) \
-		for (i = 0; i < BITS_TO_LONGS(max); i++) \
-			if ((id->bit[i] & dev->bit[i]) != id->bit[i]) \
-				break; \
-		if (i != BITS_TO_LONGS(max)) \
-			continue;
-
 static const struct input_device_id *input_match_device(struct input_handler *handler,
 							struct input_dev *dev)
 {
 	const struct input_device_id *id;
-	int i;
 
 	for (id = handler->id_table; id->flags || id->driver_info; id++) {
 
@@ -875,15 +867,32 @@ static const struct input_device_id *input_match_device(struct input_handler *ha
 			if (id->version != dev->id.version)
 				continue;
 
-		MATCH_BIT(evbit,  EV_MAX);
-		MATCH_BIT(keybit, KEY_MAX);
-		MATCH_BIT(relbit, REL_MAX);
-		MATCH_BIT(absbit, ABS_MAX);
-		MATCH_BIT(mscbit, MSC_MAX);
-		MATCH_BIT(ledbit, LED_MAX);
-		MATCH_BIT(sndbit, SND_MAX);
-		MATCH_BIT(ffbit,  FF_MAX);
-		MATCH_BIT(swbit,  SW_MAX);
+		if (!bitmap_subset(id->evbit, dev->evbit, EV_MAX))
+			continue;
+
+		if (!bitmap_subset(id->keybit, dev->keybit, KEY_MAX))
+			continue;
+
+		if (!bitmap_subset(id->relbit, dev->relbit, REL_MAX))
+			continue;
+
+		if (!bitmap_subset(id->absbit, dev->absbit, ABS_MAX))
+			continue;
+
+		if (!bitmap_subset(id->mscbit, dev->mscbit, MSC_MAX))
+			continue;
+
+		if (!bitmap_subset(id->ledbit, dev->ledbit, LED_MAX))
+			continue;
+
+		if (!bitmap_subset(id->sndbit, dev->sndbit, SND_MAX))
+			continue;
+
+		if (!bitmap_subset(id->ffbit, dev->ffbit, FF_MAX))
+			continue;
+
+		if (!bitmap_subset(id->swbit, dev->swbit, SW_MAX))
+			continue;
 
 		if (!handler->match || handler->match(handler, dev))
 			return id;

[PATCH 1/8] fs/namespace.c: introduce helper function path_unmounted()

[Yan Hong]

In fs/namespace.c, we verify whether a path is a mountpoint by
comparing path->dentry and path->mnt->mnt_root. Introduce
path_unmounted() to make code more readable.

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/namespace.c |   19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)

diff --git a/fs/namespace.c b/fs/namespace.c
index 4d31f73..3fdc239 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1234,6 +1234,11 @@ static int do_umount(struct mount *mnt, int flags)
 	return retval;
 }
 
+static inline bool path_unmounted(struct path *path)
+{
+	return path->dentry != path->mnt->mnt_root;
+}
+
 /*
  * Now umount can handle mount points as well as block devices.
  * This is important for filesystems which use unnamed block devices.
@@ -1260,7 +1265,7 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags)
 		goto out;
 	mnt = real_mount(path.mnt);
 	retval = -EINVAL;
-	if (path.dentry != path.mnt->mnt_root)
+	if (path_unmounted(&path))
 		goto dput_and_out;
 	if (!check_mnt(mnt))
 		goto dput_and_out;
@@ -1613,7 +1618,7 @@ static int do_change_type(struct path *path, int flag)
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	if (path->dentry != path->mnt->mnt_root)
+	if (path_unmounted(path))
 		return -EINVAL;
 
 	type = flags_to_propagation_type(flag);
@@ -1727,7 +1732,7 @@ static int do_remount(struct path *path, int flags, int mnt_flags,
 	if (!check_mnt(mnt))
 		return -EINVAL;
 
-	if (path->dentry != path->mnt->mnt_root)
+	if (path_unmounted(path))
 		return -EINVAL;
 
 	err = security_sb_remount(sb, data);
@@ -1793,7 +1798,7 @@ static int do_move_mount(struct path *path, char *old_name)
 		goto out1;
 
 	err = -EINVAL;
-	if (old_path.dentry != old_path.mnt->mnt_root)
+	if (path_unmounted(&old_path))
 		goto out1;
 
 	if (!mnt_has_parent(old))
@@ -1892,7 +1897,7 @@ static int do_add_mount(struct mount *newmnt, struct path *path, int mnt_flags)
 	/* Refuse the same filesystem on the same mount point */
 	err = -EBUSY;
 	if (path->mnt->mnt_sb == newmnt->mnt.mnt_sb &&
-	    path->mnt->mnt_root == path->dentry)
+	    !path_unmounted(path))
 		goto unlock;
 
 	err = -EINVAL;
@@ -2534,11 +2539,11 @@ SYSCALL_DEFINE2(pivot_root, const char __user *, new_root,
 	    old.mnt == root.mnt)
 		goto out4; /* loop, on the same file system  */
 	error = -EINVAL;
-	if (root.mnt->mnt_root != root.dentry)
+	if (path_unmounted(&root))
 		goto out4; /* not a mountpoint */
 	if (!mnt_has_parent(root_mnt))
 		goto out4; /* not attached */
-	if (new.mnt->mnt_root != new.dentry)
+	if (path_unmounted(&new))
 		goto out4; /* not a mountpoint */
 	if (!mnt_has_parent(new_mnt))
 		goto out4; /* not attached */
-- 
1.7.9.5

[PATCH 2/8] fs/namespace.c: remove unused macro MNT_WRITER_UNDERFLOW_LIMIT

[Yan Hong]

Its users have gone.

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/namespace.c |    1 -
 1 file changed, 1 deletion(-)

diff --git a/fs/namespace.c b/fs/namespace.c
index 3fdc239..dd969f8 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -59,7 +59,6 @@ static inline unsigned long hash(struct vfsmount *mnt, struct dentry *dentry)
 	return tmp & (HASH_SIZE - 1);
 }
 
-#define MNT_WRITER_UNDERFLOW_LIMIT -(1<<16)
 
 /*
  * allocation is serialized by namespace_sem, but we need the spinlock to
-- 
1.7.9.5

[PATCH 3/8] fs/namespace.c: trivial code clean

[Yan Hong]

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/namespace.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/namespace.c b/fs/namespace.c
index dd969f8..bbe9014 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -763,7 +763,7 @@ vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void
 
 	mnt->mnt.mnt_root = root;
 	mnt->mnt.mnt_sb = root->d_sb;
-	mnt->mnt_mountpoint = mnt->mnt.mnt_root;
+	mnt->mnt_mountpoint = root;
 	mnt->mnt_parent = mnt;
 	br_write_lock(&vfsmount_lock);
 	list_add_tail(&mnt->mnt_instance, &root->d_sb->s_mounts);
@@ -798,7 +798,7 @@ static struct mount *clone_mnt(struct mount *old, struct dentry *root,
 	atomic_inc(&sb->s_active);
 	mnt->mnt.mnt_sb = sb;
 	mnt->mnt.mnt_root = dget(root);
-	mnt->mnt_mountpoint = mnt->mnt.mnt_root;
+	mnt->mnt_mountpoint = root;
 	mnt->mnt_parent = mnt;
 	br_write_lock(&vfsmount_lock);
 	list_add_tail(&mnt->mnt_instance, &sb->s_mounts);
-- 
1.7.9.5

[PATCH 4/8] fs/namespace.c: check permission early in sys_[u]mount

[Yan Hong]

We have several branches in sys_mount, each of them will check
CAP_SYS_ADMIN capability seperately. Do this check at the beginning
of sys_mount.

Also check permission as early as possible in sys_umount.

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/namespace.c |   26 +++++++-------------------
 1 file changed, 7 insertions(+), 19 deletions(-)

diff --git a/fs/namespace.c b/fs/namespace.c
index bbe9014..ca2b6e9 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1256,6 +1256,9 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags)
 	if (flags & ~(MNT_FORCE | MNT_DETACH | MNT_EXPIRE | UMOUNT_NOFOLLOW))
 		return -EINVAL;
 
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
 	if (!(flags & UMOUNT_NOFOLLOW))
 		lookup_flags |= LOOKUP_FOLLOW;
 
@@ -1269,10 +1272,6 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags)
 	if (!check_mnt(mnt))
 		goto dput_and_out;
 
-	retval = -EPERM;
-	if (!capable(CAP_SYS_ADMIN))
-		goto dput_and_out;
-
 	retval = do_umount(mnt, flags);
 dput_and_out:
 	/* we mustn't call path_put() as that would clear mnt_expiry_mark */
@@ -1296,9 +1295,7 @@ SYSCALL_DEFINE1(oldumount, char __user *, name)
 
 static int mount_is_safe(struct path *path)
 {
-	if (capable(CAP_SYS_ADMIN))
-		return 0;
-	return -EPERM;
+	return 0;
 #ifdef notyet
 	if (S_ISLNK(path->dentry->d_inode->i_mode))
 		return -EPERM;
@@ -1614,9 +1611,6 @@ static int do_change_type(struct path *path, int flag)
 	int type;
 	int err = 0;
 
-	if (!capable(CAP_SYS_ADMIN))
-		return -EPERM;
-
 	if (path_unmounted(path))
 		return -EINVAL;
 
@@ -1725,9 +1719,6 @@ static int do_remount(struct path *path, int flags, int mnt_flags,
 	struct super_block *sb = path->mnt->mnt_sb;
 	struct mount *mnt = real_mount(path->mnt);
 
-	if (!capable(CAP_SYS_ADMIN))
-		return -EPERM;
-
 	if (!check_mnt(mnt))
 		return -EINVAL;
 
@@ -1774,8 +1765,6 @@ static int do_move_mount(struct path *path, char *old_name)
 	struct mount *p;
 	struct mount *old;
 	int err = 0;
-	if (!capable(CAP_SYS_ADMIN))
-		return -EPERM;
 	if (!old_name || !*old_name)
 		return -EINVAL;
 	err = kern_path(old_name, LOOKUP_FOLLOW, &old_path);
@@ -1924,10 +1913,6 @@ static int do_new_mount(struct path *path, char *type, int flags,
 	if (!type)
 		return -EINVAL;
 
-	/* we need capabilities... */
-	if (!capable(CAP_SYS_ADMIN))
-		return -EPERM;
-
 	mnt = do_kern_mount(type, flags, name, data);
 	if (IS_ERR(mnt))
 		return PTR_ERR(mnt);
@@ -2410,6 +2395,9 @@ SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name,
 	char *kernel_dev;
 	unsigned long data_page;
 
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
 	ret = copy_mount_string(type, &kernel_type);
 	if (ret < 0)
 		goto out_type;
-- 
1.7.9.5

[PATCH 5/8] fs/namei.c: introduce macro AT_FDINV

[Yan Hong]

File descriptor is irrelevent when LOOKUP_ROOT is set.
Introduce AT_FDINV to avoid using hard coded value or reusing
existing macro.

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/namei.c            |   11 ++++++-----
 include/linux/fcntl.h |    2 ++
 2 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index c5b85b3..5ffd97d 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2034,8 +2034,8 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt,
 	nd.root.dentry = dentry;
 	nd.root.mnt = mnt;
 	BUG_ON(flags & LOOKUP_PARENT);
-	/* the first argument of do_path_lookup() is ignored with LOOKUP_ROOT */
-	err = do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, &nd);
+
+	err = do_path_lookup(AT_FDINV, name, flags | LOOKUP_ROOT, &nd);
 	if (!err)
 		*path = nd.path;
 	return err;
@@ -2969,11 +2969,12 @@ struct file *do_file_open_root(struct dentry *dentry, struct vfsmount *mnt,
 	if (dentry->d_inode->i_op->follow_link && op->intent & LOOKUP_OPEN)
 		return ERR_PTR(-ELOOP);
 
-	file = path_openat(-1, name, &nd, op, flags | LOOKUP_RCU);
+	file = path_openat(AT_FDINV, name, &nd, op, flags | LOOKUP_RCU);
 	if (unlikely(file == ERR_PTR(-ECHILD)))
-		file = path_openat(-1, name, &nd, op, flags);
+		file = path_openat(AT_FDINV, name, &nd, op, flags);
 	if (unlikely(file == ERR_PTR(-ESTALE)))
-		file = path_openat(-1, name, &nd, op, flags | LOOKUP_REVAL);
+		file = path_openat(AT_FDINV, name, &nd, op,
+						flags | LOOKUP_REVAL);
 	return file;
 }
 
diff --git a/include/linux/fcntl.h b/include/linux/fcntl.h
index f550f89..abd94fe 100644
--- a/include/linux/fcntl.h
+++ b/include/linux/fcntl.h
@@ -41,6 +41,8 @@
 #define AT_FDCWD		-100    /* Special value used to indicate
                                            openat should use the current
                                            working directory. */
+#define AT_FDINV		-200	/* Special value used when LOOKUP_ROOT
+					   is set. */
 #define AT_SYMLINK_NOFOLLOW	0x100   /* Do not follow symbolic links.  */
 #define AT_REMOVEDIR		0x200   /* Remove directory instead of
                                            unlinking file.  */
-- 
1.7.9.5

[PATCH 6/8] fs/inode.c: call alloc_inode() in new_inode() directly

[Yan Hong]

This saves us a list head initialization.

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/inode.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index ac8d904..3a2cd41 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -921,9 +921,11 @@ struct inode *new_inode(struct super_block *sb)
 
 	spin_lock_prefetch(&inode_sb_list_lock);
 
-	inode = new_inode_pseudo(sb);
-	if (inode)
+	inode = alloc_inode(sb);
+	if (inode) {
+		inode->i_state = 0;
 		inode_sb_list_add(inode);
+	}
 	return inode;
 }
 EXPORT_SYMBOL(new_inode);
-- 
1.7.9.5

[PATCH 7/8] fs/inode.c: remove outstanding spin lock prefetch

[Yan Hong]

Do we have particular reason to do this here?

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/inode.c |    3 ---
 1 file changed, 3 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index 3a2cd41..e89d30c 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -14,7 +14,6 @@
 #include <linux/fsnotify.h>
 #include <linux/mount.h>
 #include <linux/posix_acl.h>
-#include <linux/prefetch.h>
 #include <linux/buffer_head.h> /* for inode_has_buffers */
 #include <linux/ratelimit.h>
 #include "internal.h"
@@ -919,8 +918,6 @@ struct inode *new_inode(struct super_block *sb)
 {
 	struct inode *inode;
 
-	spin_lock_prefetch(&inode_sb_list_lock);
-
 	inode = alloc_inode(sb);
 	if (inode) {
 		inode->i_state = 0;
-- 
1.7.9.5

[PATCH 8/8] vfs: misc comment clean

[Yan Hong]

Mostly fix comments which reference inexist locks or parameters.

Signed-off-by: Yan Hong <clouds.yan@gmail.com>
---
 fs/dcache.c    |   12 ++++--------
 fs/inode.c     |    4 ++--
 fs/namei.c     |    2 +-
 fs/namespace.c |    9 +++------
 4 files changed, 10 insertions(+), 17 deletions(-)

diff --git a/fs/dcache.c b/fs/dcache.c
index 8086636..2d97518 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -44,8 +44,6 @@
  * Usage:
  * dcache->d_inode->i_lock protects:
  *   - i_dentry, d_alias, d_inode of aliases
- * dcache_hash_bucket lock protects:
- *   - the dcache hash table
  * s_anon bl list spinlock protects:
  *   - the s_anon list (see __d_drop)
  * dcache_lru_lock protects:
@@ -64,7 +62,6 @@
  * dentry->d_inode->i_lock
  *   dentry->d_lock
  *     dcache_lru_lock
- *     dcache_hash_bucket lock
  *     s_anon lock
  *
  * If there is an ancestor relationship:
@@ -145,10 +142,9 @@ int proc_nr_dentry(ctl_table *table, int write, void __user *buffer,
 
 #include <asm/word-at-a-time.h>
 /*
- * NOTE! 'cs' and 'scount' come from a dentry, so it has a
- * aligned allocation for this particular component. We don't
- * strictly need the load_unaligned_zeropad() safety, but it
- * doesn't hurt either.
+ * NOTE! 'cs' comes from a dentry, so it has an aligned allocation
+ * for this particular component. We don't strictly need the
+ * load_unaligned_zeropad() safety, but it doesn't hurt either.
  *
  * In contrast, 'ct' and 'tcount' can be from a pathname, and do
  * need the careful unaligned handling.
@@ -304,7 +300,7 @@ static void dentry_unlink_inode(struct dentry * dentry)
 }
 
 /*
- * dentry_lru_(add|del|prune|move_tail) must be called with d_lock held.
+ * dentry_lru_(add|del|prune|move_list) must be called with d_lock held.
  */
 static void dentry_lru_add(struct dentry *dentry)
 {
diff --git a/fs/inode.c b/fs/inode.c
index e89d30c..46d4f16 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -676,7 +676,7 @@ static int can_unuse(struct inode *inode)
  * Walk the superblock inode LRU for freeable inodes and attempt to free them.
  * This is called from the superblock shrinker function with a number of inodes
  * to trim from the LRU. Inodes to be freed are moved to a temporary list and
- * then are freed outside inode_lock by dispose_list().
+ * then are freed outside s_inode_lru_lock by dispose_list().
  *
  * Any inodes which are pinned purely because of attached pagecache have their
  * pagecache removed.  If the inode has metadata buffers attached to
@@ -777,7 +777,7 @@ void prune_icache_sb(struct super_block *sb, int nr_to_scan)
 
 static void __wait_on_freeing_inode(struct inode *inode);
 /*
- * Called with the inode lock held.
+ * Called with the inode_hash_lock held.
  */
 static struct inode *find_inode(struct super_block *sb,
 				struct hlist_head *head,
diff --git a/fs/namei.c b/fs/namei.c
index 5ffd97d..7b1f714 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -421,7 +421,7 @@ EXPORT_SYMBOL(path_put);
  * Path walking has 2 modes, rcu-walk and ref-walk (see
  * Documentation/filesystems/path-lookup.txt).  In situations when we can't
  * continue in RCU mode, we attempt to drop out of rcu-walk mode and grab
- * normal reference counts on dentries and vfsmounts to transition to rcu-walk
+ * normal reference counts on dentries and vfsmounts to transition to ref-walk
  * mode.  Refcounts are grabbed at the last known good point before rcu-walk
  * got stuck, so ref-walk may continue from there. If this is not successful
  * (eg. a seqcount has changed), then failure is returned and it's up to caller
diff --git a/fs/namespace.c b/fs/namespace.c
index ca2b6e9..b7477ee 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1239,9 +1239,6 @@ static inline bool path_unmounted(struct path *path)
 }
 
 /*
- * Now umount can handle mount points as well as block devices.
- * This is important for filesystems which use unnamed block devices.
- *
  * We now support a flag for forced unmount like the other 'big iron'
  * unixes. Our API is identical to OSF/1 to avoid making a mess of AMD
  */
@@ -1431,9 +1428,9 @@ static int invent_group_ids(struct mount *mnt, bool recurse)
 }
 
 /*
- *  @source_mnt : mount tree to be attached
- *  @nd         : place the mount tree @source_mnt is attached
- *  @parent_nd  : if non-null, detach the source_mnt from its parent and
+ *  @source_mnt  : mount tree to be attached
+ *  @path        : place the mount tree @source_mnt is attached
+ *  @parent_path : if non-null, detach the @source_mnt from its parent and
  *  		   store the parent mount and mountpoint dentry.
  *  		   (done when source_mnt is moved)
  *
-- 
1.7.9.5

[PATCH] Staging: android: fixed const coding style issue in binder.c

[Patrik Karlin]

This is a patch to add const to the lists of strings in binder.c
as suggested by checkpatch.pl

Signed-off-by: Patrik Kårlin <patrik.karlin@gmail.com>
---
 drivers/staging/android/binder.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/staging/android/binder.c b/drivers/staging/android/binder.c
index 2d12e8a..538ebe2 100644
--- a/drivers/staging/android/binder.c
+++ b/drivers/staging/android/binder.c
@@ -3227,7 +3227,7 @@ static void print_binder_proc(struct seq_file *m,
 		m->count = start_pos;
 }
 
-static const char *binder_return_strings[] = {
+static const char * const binder_return_strings[] = {
 	"BR_ERROR",
 	"BR_OK",
 	"BR_TRANSACTION",
@@ -3248,7 +3248,7 @@ static const char *binder_return_strings[] = {
 	"BR_FAILED_REPLY"
 };
 
-static const char *binder_command_strings[] = {
+static const char * const binder_command_strings[] = {
 	"BC_TRANSACTION",
 	"BC_REPLY",
 	"BC_ACQUIRE_RESULT",
@@ -3268,7 +3268,7 @@ static const char *binder_command_strings[] = {
 	"BC_DEAD_BINDER_DONE"
 };
 
-static const char *binder_objstat_strings[] = {
+static const char * const binder_objstat_strings[] = {
 	"proc",
 	"thread",
 	"node",
-- 
1.7.10.4

Re: [PATCH] Staging: android: fixed const coding style issue in binder.c

[Greg KH]

On Mon, Jan 07, 2013 at 07:11:52PM +0100, Patrik Karlin wrote:
> This is a patch to add const to the lists of strings in binder.c
> as suggested by checkpatch.pl
> 
> Signed-off-by: Patrik Kårlin <patrik.karlin@gmail.com>
> ---
>  drivers/staging/android/binder.c |    6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)

Sorry, I just applied a patch from someone else that did this same
thing, that was sent to me a few weeks ago.

greg k-h

[PATCH v2] SUNRPC: Allow one callback request to be received from two sk_buff

[shaobingqing]

In current code, there only one struct rpc_rqst is prealloced. If one
callback request is received from two sk_buff, the xprt_alloc_bc_request
would be execute two times with the same transport->xid. The first time
xprt_alloc_bc_request will alloc one struct rpc_rqst and the TCP_RCV_COPY_DATA
bit of transport->tcp_flags will not be cleared. The second time
xprt_alloc_bc_request could not alloc struct rpc_rqst any more and NULL
pointer will be returned, then xprt_force_disconnect occur. I think one
callback request can be allowed to be received from two sk_buff.

Signed-off-by: shaobingqing <shaobingqing@bwstor.com.cn>
---
 include/linux/sunrpc/xprt.h |    1 +
 net/sunrpc/xprt.c           |    1 +
 net/sunrpc/xprtsock.c       |   13 ++++++++++++-
 3 files changed, 14 insertions(+), 1 deletions(-)

diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index cec7b9b..82bfe01 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -211,6 +211,7 @@ struct rpc_xprt {
 						 * items */
 	struct list_head	bc_pa_list;	/* List of preallocated
 						 * backchannel rpc_rqst's */
+	struct rpc_rqst	*req_first;
 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
 	struct list_head	recv;
 
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 095363e..93ad8bc 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -1256,6 +1256,7 @@ static void xprt_init(struct rpc_xprt *xprt, struct net *net)
 #if defined(CONFIG_SUNRPC_BACKCHANNEL)
 	spin_lock_init(&xprt->bc_pa_lock);
 	INIT_LIST_HEAD(&xprt->bc_pa_list);
+	xprt->req_first = NULL;
 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 	xprt->last_used = jiffies;
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index ee03d35..c43dca4 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -1272,7 +1272,16 @@ static inline int xs_tcp_read_callback(struct rpc_xprt *xprt,
 				container_of(xprt, struct sock_xprt, xprt);
 	struct rpc_rqst *req;
 
-	req = xprt_alloc_bc_request(xprt);
+	if (xprt->req_first != NULL &&
+			xprt->req_first->rq_xid == transport->tcp_xid) {
+		req = xprt->req_first;
+	} else if (xprt->req_first != NULL &&
+			xprt->req_first->rq_xid != transport->tcp_xid) {
+		xprt_free_bc_request(xprt);
+		req = xprt_alloc_bc_request(xprt);
+	} else {
+		req = xprt_alloc_bc_request(xprt);
+	}
 	if (req == NULL) {
 		printk(KERN_WARNING "Callback slot table overflowed\n");
 		xprt_force_disconnect(xprt);
@@ -1297,6 +1306,8 @@ static inline int xs_tcp_read_callback(struct rpc_xprt *xprt,
 		list_add(&req->rq_bc_list, &bc_serv->sv_cb_list);
 		spin_unlock(&bc_serv->sv_cb_lock);
 		wake_up(&bc_serv->sv_cb_waitq);
+	} else {
+		xprt->req_first = req;
 	}
 
 	req->rq_private_buf.len = transport->tcp_copied;
-- 
1.7.4.2

Re: [PATCH v2] SUNRPC: Allow one callback request to be received from two sk_buff

[Sergei Shtylyov]

Hello.

On 08-02-2014 6:29, shaobingqing wrote:

> In current code, there only one struct rpc_rqst is prealloced. If one
> callback request is received from two sk_buff, the xprt_alloc_bc_request
> would be execute two times with the same transport->xid. The first time
> xprt_alloc_bc_request will alloc one struct rpc_rqst and the TCP_RCV_COPY_DATA
> bit of transport->tcp_flags will not be cleared. The second time
> xprt_alloc_bc_request could not alloc struct rpc_rqst any more and NULL
> pointer will be returned, then xprt_force_disconnect occur. I think one
> callback request can be allowed to be received from two sk_buff.

> Signed-off-by: shaobingqing <shaobingqing@bwstor.com.cn>
[...]

> diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
> index cec7b9b..82bfe01 100644
> --- a/include/linux/sunrpc/xprt.h
> +++ b/include/linux/sunrpc/xprt.h
> @@ -211,6 +211,7 @@ struct rpc_xprt {
>   						 * items */
>   	struct list_head	bc_pa_list;	/* List of preallocated
>   						 * backchannel rpc_rqst's */
> +	struct rpc_rqst	*req_first;

    Please align the variable name with the ones above and below.

>   #endif /* CONFIG_SUNRPC_BACKCHANNEL */
>   	struct list_head	recv;
>
[...]
> diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
> index ee03d35..c43dca4 100644
> --- a/net/sunrpc/xprtsock.c
> +++ b/net/sunrpc/xprtsock.c
> @@ -1272,7 +1272,16 @@ static inline int xs_tcp_read_callback(struct rpc_xprt *xprt,
>   				container_of(xprt, struct sock_xprt, xprt);
>   	struct rpc_rqst *req;
>
> -	req = xprt_alloc_bc_request(xprt);
> +	if (xprt->req_first != NULL &&
> +			xprt->req_first->rq_xid == transport->tcp_xid) {

    The continuation line should be aligned right under 'xprt' on the broken 
up line, according to the networking coding style.

> +		req = xprt->req_first;
> +	} else if (xprt->req_first != NULL &&
> +			xprt->req_first->rq_xid != transport->tcp_xid) {

    Same here.

WBR, Sergei

Re: [PATCH v2] SUNRPC: Allow one callback request to be received from two sk_buff

[Trond Myklebust]

On Sat, 2014-02-08 at 10:29 +0800, shaobingqing wrote:
> In current code, there only one struct rpc_rqst is prealloced. If one
> callback request is received from two sk_buff, the xprt_alloc_bc_request
> would be execute two times with the same transport->xid. The first time
> xprt_alloc_bc_request will alloc one struct rpc_rqst and the TCP_RCV_COPY_DATA
> bit of transport->tcp_flags will not be cleared. The second time
> xprt_alloc_bc_request could not alloc struct rpc_rqst any more and NULL
> pointer will be returned, then xprt_force_disconnect occur. I think one
> callback request can be allowed to be received from two sk_buff.
> 
> Signed-off-by: shaobingqing <shaobingqing@bwstor.com.cn>
> ---

There are still several problems with this patch.
For one thing, xprt->req_first is never cleared, which means that the
tests you add in xs_tcp_read_callback() could trigger even if the
request is in use by the callback server.
Furthermore, if the xprt is destroyed before the RPC callback has been
filled, then the request is leaked: it is no longer on the xprt
backchannel list, so nothing will clean it up there, and nothing ever
checks xprt->req_first, so it won't get cleaned up there either.

Instead of adding a new tracking mechanism, what say we just keep the
request on the xprt callback list until it has been completely filled.

IOW: something like the following patch:

8<-------------------------------------------------------------------
>From 2aa57b6c6c6025e17d7ca164bbda01767af662be Mon Sep 17 00:00:00 2001
From: Trond Myklebust <trond.myklebust@primarydata.com>
Date: Mon, 10 Feb 2014 11:18:39 -0500
Subject: [PATCH] SUNRPC: RPC callbacks may be split across several TCP
 segments

Since TCP is a stream protocol, our callback read code needs to take into
account the fact that RPC callbacks are not always confined to a single
TCP segment.
This patch adds support for multiple TCP segments by ensuring that we
only remove the rpc_rqst structure from the 'free backchannel requests'
list once the data has been completely received. We rely on the fact
that TCP data is ordered for the duration of the connection.

Reported-by: shaobingqing <shaobingqing@bwstor.com.cn>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
---
 include/linux/sunrpc/bc_xprt.h |  3 +-
 net/sunrpc/backchannel_rqst.c  | 92 +++++++++++++++++++++++++++++-------------
 net/sunrpc/xprtsock.c          | 28 ++++---------
 3 files changed, 73 insertions(+), 50 deletions(-)

diff --git a/include/linux/sunrpc/bc_xprt.h b/include/linux/sunrpc/bc_xprt.h
index 969c0a671dbf..2ca67b55e0fe 100644
--- a/include/linux/sunrpc/bc_xprt.h
+++ b/include/linux/sunrpc/bc_xprt.h
@@ -32,7 +32,8 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #include <linux/sunrpc/sched.h>
 
 #ifdef CONFIG_SUNRPC_BACKCHANNEL
-struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt);
+struct rpc_rqst *xprt_lookup_bc_request(struct rpc_xprt *xprt, __be32 xid);
+void xprt_complete_bc_request(struct rpc_rqst *req, uint32_t copied);
 void xprt_free_bc_request(struct rpc_rqst *req);
 int xprt_setup_backchannel(struct rpc_xprt *, unsigned int min_reqs);
 void xprt_destroy_backchannel(struct rpc_xprt *, unsigned int max_reqs);
diff --git a/net/sunrpc/backchannel_rqst.c b/net/sunrpc/backchannel_rqst.c
index 890a29912d5a..7cceeb06b939 100644
--- a/net/sunrpc/backchannel_rqst.c
+++ b/net/sunrpc/backchannel_rqst.c
@@ -210,39 +210,23 @@ out:
 }
 EXPORT_SYMBOL_GPL(xprt_destroy_backchannel);
 
-/*
- * One or more rpc_rqst structure have been preallocated during the
- * backchannel setup.  Buffer space for the send and private XDR buffers
- * has been preallocated as well.  Use xprt_alloc_bc_request to allocate
- * to this request.  Use xprt_free_bc_request to return it.
- *
- * We know that we're called in soft interrupt context, grab the spin_lock
- * since there is no need to grab the bottom half spin_lock.
- *
- * Return an available rpc_rqst, otherwise NULL if non are available.
- */
-struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt)
+static struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt, __be32 xid)
 {
-	struct rpc_rqst *req;
+	struct rpc_rqst *req = NULL;
 
 	dprintk("RPC:       allocate a backchannel request\n");
-	spin_lock(&xprt->bc_pa_lock);
-	if (!list_empty(&xprt->bc_pa_list)) {
-		req = list_first_entry(&xprt->bc_pa_list, struct rpc_rqst,
-				rq_bc_pa_list);
-		list_del(&req->rq_bc_pa_list);
-	} else {
-		req = NULL;
-	}
-	spin_unlock(&xprt->bc_pa_lock);
+	if (list_empty(&xprt->bc_pa_list))
+		goto not_found;
 
-	if (req != NULL) {
-		set_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state);
-		req->rq_reply_bytes_recvd = 0;
-		req->rq_bytes_sent = 0;
-		memcpy(&req->rq_private_buf, &req->rq_rcv_buf,
+	req = list_first_entry(&xprt->bc_pa_list, struct rpc_rqst,
+				rq_bc_pa_list);
+	req->rq_reply_bytes_recvd = 0;
+	req->rq_bytes_sent = 0;
+	memcpy(&req->rq_private_buf, &req->rq_rcv_buf,
 			sizeof(req->rq_private_buf));
-	}
+	req->rq_xid = xid;
+	req->rq_connect_cookie = xprt->connect_cookie;
+not_found:
 	dprintk("RPC:       backchannel req=%p\n", req);
 	return req;
 }
@@ -257,6 +241,7 @@ void xprt_free_bc_request(struct rpc_rqst *req)
 
 	dprintk("RPC:       free backchannel req=%p\n", req);
 
+	req->rq_connect_cookie = xprt->connect_cookie - 1;
 	smp_mb__before_clear_bit();
 	WARN_ON_ONCE(!test_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state));
 	clear_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state);
@@ -279,7 +264,56 @@ void xprt_free_bc_request(struct rpc_rqst *req)
 	 * may be reused by a new callback request.
 	 */
 	spin_lock_bh(&xprt->bc_pa_lock);
-	list_add(&req->rq_bc_pa_list, &xprt->bc_pa_list);
+	list_add_tail(&req->rq_bc_pa_list, &xprt->bc_pa_list);
 	spin_unlock_bh(&xprt->bc_pa_lock);
 }
 
+/*
+ * One or more rpc_rqst structure have been preallocated during the
+ * backchannel setup.  Buffer space for the send and private XDR buffers
+ * has been preallocated as well.  Use xprt_alloc_bc_request to allocate
+ * to this request.  Use xprt_free_bc_request to return it.
+ *
+ * We know that we're called in soft interrupt context, grab the spin_lock
+ * since there is no need to grab the bottom half spin_lock.
+ *
+ * Return an available rpc_rqst, otherwise NULL if non are available.
+ */
+struct rpc_rqst *xprt_lookup_bc_request(struct rpc_xprt *xprt, __be32 xid)
+{
+	struct rpc_rqst *req;
+
+	spin_lock(&xprt->bc_pa_lock);
+	list_for_each_entry(req, &xprt->bc_pa_list, rq_bc_pa_list) {
+		if (req->rq_connect_cookie != xprt->connect_cookie)
+			continue;
+		if (req->rq_xid == xid)
+			goto found;
+	}
+	req = xprt_alloc_bc_request(xprt, xid);
+found:
+	spin_unlock(&xprt->bc_pa_lock);
+	return req;
+}
+
+/*
+ * Add callback request to callback list.  The callback
+ * service sleeps on the sv_cb_waitq waiting for new
+ * requests.  Wake it up after adding enqueing the
+ * request.
+ */
+void xprt_complete_bc_request(struct rpc_rqst *req, uint32_t copied)
+{
+	struct rpc_xprt *xprt = req->rq_xprt;
+	struct svc_serv *bc_serv = xprt->bc_serv;
+
+	req->rq_private_buf.len = copied;
+	set_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state);
+
+	dprintk("RPC:       add callback request to list\n");
+	spin_lock(&bc_serv->sv_cb_lock);
+	list_move(&req->rq_bc_list, &bc_serv->sv_cb_list);
+	wake_up(&bc_serv->sv_cb_waitq);
+	spin_unlock(&bc_serv->sv_cb_lock);
+}
+
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 817a1e523969..6497c221612c 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -1302,41 +1302,29 @@ static inline int xs_tcp_read_reply(struct rpc_xprt *xprt,
  * If we're unable to obtain the rpc_rqst we schedule the closing of the
  * connection and return -1.
  */
-static inline int xs_tcp_read_callback(struct rpc_xprt *xprt,
+static int xs_tcp_read_callback(struct rpc_xprt *xprt,
 				       struct xdr_skb_reader *desc)
 {
 	struct sock_xprt *transport =
 				container_of(xprt, struct sock_xprt, xprt);
 	struct rpc_rqst *req;
 
-	req = xprt_alloc_bc_request(xprt);
+	/* Look up and lock the request corresponding to the given XID */
+	spin_lock(&xprt->transport_lock);
+	req = xprt_lookup_bc_request(xprt, transport->tcp_xid);
 	if (req == NULL) {
+		spin_unlock(&xprt->transport_lock);
 		printk(KERN_WARNING "Callback slot table overflowed\n");
 		xprt_force_disconnect(xprt);
 		return -1;
 	}
 
-	req->rq_xid = transport->tcp_xid;
 	dprintk("RPC:       read callback  XID %08x\n", ntohl(req->rq_xid));
 	xs_tcp_read_common(xprt, desc, req);
 
-	if (!(transport->tcp_flags & TCP_RCV_COPY_DATA)) {
-		struct svc_serv *bc_serv = xprt->bc_serv;
-
-		/*
-		 * Add callback request to callback list.  The callback
-		 * service sleeps on the sv_cb_waitq waiting for new
-		 * requests.  Wake it up after adding enqueing the
-		 * request.
-		 */
-		dprintk("RPC:       add callback request to list\n");
-		spin_lock(&bc_serv->sv_cb_lock);
-		list_add(&req->rq_bc_list, &bc_serv->sv_cb_list);
-		spin_unlock(&bc_serv->sv_cb_lock);
-		wake_up(&bc_serv->sv_cb_waitq);
-	}
-
-	req->rq_private_buf.len = transport->tcp_copied;
+	if (!(transport->tcp_flags & TCP_RCV_COPY_DATA))
+		xprt_complete_bc_request(req, transport->tcp_copied);
+	spin_unlock(&xprt->transport_lock);
 
 	return 0;
 }
-- 
1.8.5.3


-- 
Trond Myklebust
Linux NFS client maintainer, PrimaryData
trond.myklebust@primarydata.com