Merge branch 'x86/irq' into x86/devel

Conflicts:

	arch/x86/kernel/i8259.c
	arch/x86/kernel/irqinit_64.c

Signed-off-by: Ingo Molnar <mingo@elte.hu>

10 files changed, 274 insertions, 184 deletions

diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
index 33c5216fd3e..ff1a7b49a46 100644
--- a/arch/x86/kernel/acpi/boot.c
+++ b/arch/x86/kernel/acpi/boot.c
@@ -514,8 +514,6 @@ int acpi_register_gsi(u32 gsi, int triggering, int polarity)
 	 * Make sure all (legacy) PCI IRQs are set as level-triggered.
 	 */
 	if (acpi_irq_model == ACPI_IRQ_MODEL_PIC) {
-		extern void eisa_set_level_irq(unsigned int irq);
-
 		if (triggering == ACPI_LEVEL_SENSITIVE)
 			eisa_set_level_irq(gsi);
 	}
diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c
index 45d8da405ad..ce4538ebb7f 100644
--- a/arch/x86/kernel/apic_32.c
+++ b/arch/x86/kernel/apic_32.c
@@ -70,6 +70,10 @@ static int local_apic_timer_disabled;
 int local_apic_timer_c2_ok;
 EXPORT_SYMBOL_GPL(local_apic_timer_c2_ok);
 
+int first_system_vector = 0xfe;
+
+char system_vectors[NR_VECTORS] = { [0 ... NR_VECTORS-1] = SYS_VECTOR_FREE};
+
 /*
  * Debug level, exported for io_apic.c
  */
@@ -1351,13 +1355,13 @@ void __init smp_intr_init(void)
 	 * The reschedule interrupt is a CPU-to-CPU reschedule-helper
 	 * IPI, driven by wakeup.
 	 */
-	set_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt);
+	alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt);
 
 	/* IPI for invalidation */
-	set_intr_gate(INVALIDATE_TLB_VECTOR, invalidate_interrupt);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR, invalidate_interrupt);
 
 	/* IPI for generic function call */
-	set_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt);
+	alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt);
 }
 #endif
 
@@ -1370,15 +1374,15 @@ void __init apic_intr_init(void)
 	smp_intr_init();
 #endif
 	/* self generated IPI for local APIC timer */
-	set_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt);
+	alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt);
 
 	/* IPI vectors for APIC spurious and error interrupts */
-	set_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt);
-	set_intr_gate(ERROR_APIC_VECTOR, error_interrupt);
+	alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt);
+	alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt);
 
 	/* thermal monitor LVT interrupt */
 #ifdef CONFIG_X86_MCE_P4THERMAL
-	set_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt);
+	alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt);
 #endif
 }
 
diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index c778e4fa55a..159a1c76d2b 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -51,7 +51,7 @@
 #include <asm/percpu.h>
 #include <asm/dwarf2.h>
 #include <asm/processor-flags.h>
-#include "irq_vectors.h"
+#include <asm/irq_vectors.h>
 
 /*
  * We use macros for low-level operations which need to be overridden
diff --git a/arch/x86/kernel/genx2apic_uv_x.c b/arch/x86/kernel/genx2apic_uv_x.c
index ebf13908a74..45e84acca8a 100644
--- a/arch/x86/kernel/genx2apic_uv_x.c
+++ b/arch/x86/kernel/genx2apic_uv_x.c
@@ -5,7 +5,7 @@
  *
  * SGI UV APIC functions (note: not an Intel compatible APIC)
  *
- * Copyright (C) 2007 Silicon Graphics, Inc. All rights reserved.
+ * Copyright (C) 2007-2008 Silicon Graphics, Inc. All rights reserved.
  */
 
 #include <linux/threads.h>
@@ -55,37 +55,37 @@ static cpumask_t uv_vector_allocation_domain(int cpu)
 int uv_wakeup_secondary(int phys_apicid, unsigned int start_rip)
 {
 	unsigned long val;
-	int nasid;
+	int pnode;
 
-	nasid = uv_apicid_to_nasid(phys_apicid);
+	pnode = uv_apicid_to_pnode(phys_apicid);
 	val = (1UL << UVH_IPI_INT_SEND_SHFT) |
 	    (phys_apicid << UVH_IPI_INT_APIC_ID_SHFT) |
 	    (((long)start_rip << UVH_IPI_INT_VECTOR_SHFT) >> 12) |
 	    APIC_DM_INIT;
-	uv_write_global_mmr64(nasid, UVH_IPI_INT, val);
+	uv_write_global_mmr64(pnode, UVH_IPI_INT, val);
 	mdelay(10);
 
 	val = (1UL << UVH_IPI_INT_SEND_SHFT) |
 	    (phys_apicid << UVH_IPI_INT_APIC_ID_SHFT) |
 	    (((long)start_rip << UVH_IPI_INT_VECTOR_SHFT) >> 12) |
 	    APIC_DM_STARTUP;
-	uv_write_global_mmr64(nasid, UVH_IPI_INT, val);
+	uv_write_global_mmr64(pnode, UVH_IPI_INT, val);
 	return 0;
 }
 
 static void uv_send_IPI_one(int cpu, int vector)
 {
 	unsigned long val, apicid, lapicid;
-	int nasid;
+	int pnode;
 
 	apicid = per_cpu(x86_cpu_to_apicid, cpu); /* ZZZ - cache node-local ? */
 	lapicid = apicid & 0x3f;		/* ZZZ macro needed */
-	nasid = uv_apicid_to_nasid(apicid);
+	pnode = uv_apicid_to_pnode(apicid);
 	val =
 	    (1UL << UVH_IPI_INT_SEND_SHFT) | (lapicid <<
 					      UVH_IPI_INT_APIC_ID_SHFT) |
 	    (vector << UVH_IPI_INT_VECTOR_SHFT);
-	uv_write_global_mmr64(nasid, UVH_IPI_INT, val);
+	uv_write_global_mmr64(pnode, UVH_IPI_INT, val);
 }
 
 static void uv_send_IPI_mask(cpumask_t mask, int vector)
@@ -159,39 +159,81 @@ struct genapic apic_x2apic_uv_x = {
 	.phys_pkg_id = phys_pkg_id,	/* Fixme ZZZ */
 };
 
-static __cpuinit void set_x2apic_extra_bits(int nasid)
+static __cpuinit void set_x2apic_extra_bits(int pnode)
 {
-	__get_cpu_var(x2apic_extra_bits) = ((nasid >> 1) << 6);
+	__get_cpu_var(x2apic_extra_bits) = (pnode << 6);
 }
 
 /*
  * Called on boot cpu.
  */
+static __init int boot_pnode_to_blade(int pnode)
+{
+	int blade;
+
+	for (blade = 0; blade < uv_num_possible_blades(); blade++)
+		if (pnode == uv_blade_info[blade].pnode)
+			return blade;
+	BUG();
+}
+
+struct redir_addr {
+	unsigned long redirect;
+	unsigned long alias;
+};
+
+#define DEST_SHIFT UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR_DEST_BASE_SHFT
+
+static __initdata struct redir_addr redir_addrs[] = {
+	{UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR, UVH_SI_ALIAS0_OVERLAY_CONFIG},
+	{UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR, UVH_SI_ALIAS1_OVERLAY_CONFIG},
+	{UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR, UVH_SI_ALIAS2_OVERLAY_CONFIG},
+};
+
+static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size)
+{
+	union uvh_si_alias0_overlay_config_u alias;
+	union uvh_rh_gam_alias210_redirect_config_2_mmr_u redirect;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(redir_addrs); i++) {
+		alias.v = uv_read_local_mmr(redir_addrs[i].alias);
+		if (alias.s.base == 0) {
+			*size = (1UL << alias.s.m_alias);
+			redirect.v = uv_read_local_mmr(redir_addrs[i].redirect);
+			*base = (unsigned long)redirect.s.dest_base << DEST_SHIFT;
+			return;
+		}
+	}
+	BUG();
+}
+
 static __init void uv_system_init(void)
 {
 	union uvh_si_addr_map_config_u m_n_config;
-	int bytes, nid, cpu, lcpu, nasid, last_nasid, blade;
-	unsigned long mmr_base;
+	union uvh_node_id_u node_id;
+	unsigned long gnode_upper, lowmem_redir_base, lowmem_redir_size;
+	int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val;
+	unsigned long mmr_base, present;
 
 	m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG);
+	m_val = m_n_config.s.m_skt;
+	n_val = m_n_config.s.n_skt;
 	mmr_base =
 	    uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) &
 	    ~UV_MMR_ENABLE;
 	printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base);
 
-	last_nasid = -1;
-	for_each_possible_cpu(cpu) {
-		nid = cpu_to_node(cpu);
-		nasid = uv_apicid_to_nasid(per_cpu(x86_cpu_to_apicid, cpu));
-		if (nasid != last_nasid)
-			uv_possible_blades++;
-		last_nasid = nasid;
-	}
+	for(i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++)
+		uv_possible_blades +=
+		  hweight64(uv_read_local_mmr( UVH_NODE_PRESENT_TABLE + i * 8));
 	printk(KERN_DEBUG "UV: Found %d blades\n", uv_num_possible_blades());
 
 	bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades();
 	uv_blade_info = alloc_bootmem_pages(bytes);
 
+	get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size);
+
 	bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes();
 	uv_node_to_blade = alloc_bootmem_pages(bytes);
 	memset(uv_node_to_blade, 255, bytes);
@@ -200,43 +242,56 @@ static __init void uv_system_init(void)
 	uv_cpu_to_blade = alloc_bootmem_pages(bytes);
 	memset(uv_cpu_to_blade, 255, bytes);
 
-	last_nasid = -1;
-	blade = -1;
-	lcpu = -1;
-	for_each_possible_cpu(cpu) {
-		nid = cpu_to_node(cpu);
-		nasid = uv_apicid_to_nasid(per_cpu(x86_cpu_to_apicid, cpu));
-		if (nasid != last_nasid) {
-			blade++;
-			lcpu = -1;
-			uv_blade_info[blade].nr_posible_cpus = 0;
+	blade = 0;
+	for (i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) {
+		present = uv_read_local_mmr(UVH_NODE_PRESENT_TABLE + i * 8);
+		for (j = 0; j < 64; j++) {
+			if (!test_bit(j, &present))
+				continue;
+			uv_blade_info[blade].pnode = (i * 64 + j);
+			uv_blade_info[blade].nr_possible_cpus = 0;
 			uv_blade_info[blade].nr_online_cpus = 0;
+			blade++;
 		}
-		last_nasid = nasid;
-		lcpu++;
+	}
 
-		uv_cpu_hub_info(cpu)->m_val = m_n_config.s.m_skt;
-		uv_cpu_hub_info(cpu)->n_val = m_n_config.s.n_skt;
+	node_id.v = uv_read_local_mmr(UVH_NODE_ID);
+	gnode_upper = (((unsigned long)node_id.s.node_id) &
+		       ~((1 << n_val) - 1)) << m_val;
+
+	for_each_present_cpu(cpu) {
+		nid = cpu_to_node(cpu);
+		pnode = uv_apicid_to_pnode(per_cpu(x86_cpu_to_apicid, cpu));
+		blade = boot_pnode_to_blade(pnode);
+		lcpu = uv_blade_info[blade].nr_possible_cpus;
+		uv_blade_info[blade].nr_possible_cpus++;
+
+		uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base;
+		uv_cpu_hub_info(cpu)->lowmem_remap_top =
+					lowmem_redir_base + lowmem_redir_size;
+		uv_cpu_hub_info(cpu)->m_val = m_val;
+		uv_cpu_hub_info(cpu)->n_val = m_val;
 		uv_cpu_hub_info(cpu)->numa_blade_id = blade;
 		uv_cpu_hub_info(cpu)->blade_processor_id = lcpu;
-		uv_cpu_hub_info(cpu)->local_nasid = nasid;
-		uv_cpu_hub_info(cpu)->gnode_upper =
-		    nasid & ~((1 << uv_hub_info->n_val) - 1);
+		uv_cpu_hub_info(cpu)->pnode = pnode;
+		uv_cpu_hub_info(cpu)->pnode_mask = (1 << n_val) - 1;
+		uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1;
+		uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper;
 		uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base;
 		uv_cpu_hub_info(cpu)->coherency_domain_number = 0;/* ZZZ */
-		uv_blade_info[blade].nasid = nasid;
-		uv_blade_info[blade].nr_posible_cpus++;
 		uv_node_to_blade[nid] = blade;
 		uv_cpu_to_blade[cpu] = blade;
 
-		printk(KERN_DEBUG "UV cpu %d, apicid 0x%x, nasid %d, nid %d\n",
-		       cpu, per_cpu(x86_cpu_to_apicid, cpu), nasid, nid);
-		printk(KERN_DEBUG "UV   lcpu %d, blade %d\n", lcpu, blade);
+		printk(KERN_DEBUG "UV cpu %d, apicid 0x%x, pnode %d, nid %d, "
+			"lcpu %d, blade %d\n",
+			cpu, per_cpu(x86_cpu_to_apicid, cpu), pnode, nid,
+			lcpu, blade);
 	}
 }
 
 /*
  * Called on each cpu to initialize the per_cpu UV data area.
+ * 	ZZZ hotplug not supported yet
  */
 void __cpuinit uv_cpu_init(void)
 {
@@ -246,5 +301,5 @@ void __cpuinit uv_cpu_init(void)
 	uv_blade_info[uv_numa_blade_id()].nr_online_cpus++;
 
 	if (get_uv_system_type() == UV_NON_UNIQUE_APIC)
-		set_x2apic_extra_bits(uv_hub_info->local_nasid);
+		set_x2apic_extra_bits(uv_hub_info->pnode);
 }
diff --git a/arch/x86/kernel/i8259.c b/arch/x86/kernel/i8259.c
index 7a0fda8f01b..dc92b49d920 100644
--- a/arch/x86/kernel/i8259.c
+++ b/arch/x86/kernel/i8259.c
@@ -297,34 +297,28 @@ void init_8259A(int auto_eoi)
 	 * outb_pic - this has to work on a wide range of PC hardware.
 	 */
 	outb_pic(0x11, PIC_MASTER_CMD);	/* ICW1: select 8259A-1 init */
-#ifndef CONFIG_X86_64
-	outb_pic(0x20 + 0, PIC_MASTER_IMR);	/* ICW2: 8259A-1 IR0-7 mapped to 0x20-0x27 */
-	outb_pic(1U << PIC_CASCADE_IR, PIC_MASTER_IMR);	/* 8259A-1 (the master) has a slave on IR2 */
-#else /* CONFIG_X86_64 */
-	/* ICW2: 8259A-1 IR0-7 mapped to 0x30-0x37 */
+
+	/* ICW2: 8259A-1 IR0-7 mapped to 0x30-0x37 on x86-64,
+	                       to 0x20-0x27 on i386 */
 	outb_pic(IRQ0_VECTOR, PIC_MASTER_IMR);
+
 	/* 8259A-1 (the master) has a slave on IR2 */
-	outb_pic(0x04, PIC_MASTER_IMR);
-#endif /* CONFIG_X86_64 */
+	outb_pic(1U << PIC_CASCADE_IR, PIC_MASTER_IMR);
+
 	if (auto_eoi)	/* master does Auto EOI */
 		outb_pic(MASTER_ICW4_DEFAULT | PIC_ICW4_AEOI, PIC_MASTER_IMR);
 	else		/* master expects normal EOI */
 		outb_pic(MASTER_ICW4_DEFAULT, PIC_MASTER_IMR);
 
 	outb_pic(0x11, PIC_SLAVE_CMD);	/* ICW1: select 8259A-2 init */
-#ifndef CONFIG_X86_64
-	outb_pic(0x20 + 8, PIC_SLAVE_IMR);	/* ICW2: 8259A-2 IR0-7 mapped to 0x28-0x2f */
-	outb_pic(PIC_CASCADE_IR, PIC_SLAVE_IMR);	/* 8259A-2 is a slave on master's IR2 */
-	outb_pic(SLAVE_ICW4_DEFAULT, PIC_SLAVE_IMR); /* (slave's support for AEOI in flat mode is to be investigated) */
-#else /* CONFIG_X86_64 */
-	/* ICW2: 8259A-2 IR0-7 mapped to 0x38-0x3f */
+
+	/* ICW2: 8259A-2 IR0-7 mapped to IRQ8_VECTOR */
 	outb_pic(IRQ8_VECTOR, PIC_SLAVE_IMR);
 	/* 8259A-2 is a slave on master's IR2 */
 	outb_pic(PIC_CASCADE_IR, PIC_SLAVE_IMR);
 	/* (slave's support for AEOI in flat mode is to be investigated) */
 	outb_pic(SLAVE_ICW4_DEFAULT, PIC_SLAVE_IMR);
 
-#endif /* CONFIG_X86_64 */
 	if (auto_eoi)
 		/*
 		 * In AEOI mode we just have to mask the interrupt
diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c
index d4f9df2b022..dac47d61d2b 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -1174,7 +1174,7 @@ static int __assign_irq_vector(int irq)
 	offset = current_offset;
 next:
 	vector += 8;
-	if (vector >= FIRST_SYSTEM_VECTOR) {
+	if (vector >= first_system_vector) {
 		offset = (offset + 1) % 8;
 		vector = FIRST_DEVICE_VECTOR + offset;
 	}
@@ -2280,7 +2280,7 @@ void __init setup_IO_APIC(void)
 	int i;
 
 	/* Reserve all the system vectors. */
-	for (i = FIRST_SYSTEM_VECTOR; i < NR_VECTORS; i++)
+	for (i = first_system_vector; i < NR_VECTORS; i++)
 		set_bit(i, used_vectors);
 
 	enable_IO_APIC();
diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c
index e5ef6030356..78a3866ab36 100644
--- a/arch/x86/kernel/io_apic_64.c
+++ b/arch/x86/kernel/io_apic_64.c
@@ -82,6 +82,10 @@ static struct irq_cfg irq_cfg[NR_IRQS] __read_mostly = {
 
 static int assign_irq_vector(int irq, cpumask_t mask);
 
+int first_system_vector = 0xfe;
+
+char system_vectors[NR_VECTORS] = { [0 ... NR_VECTORS-1] = SYS_VECTOR_FREE};
+
 #define __apicdebuginit  __init
 
 int sis_apic_bug; /* not actually supported, dummy for compile */
@@ -737,7 +741,7 @@ static int __assign_irq_vector(int irq, cpumask_t mask)
 		offset = current_offset;
 next:
 		vector += 8;
-		if (vector >= FIRST_SYSTEM_VECTOR) {
+		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;
diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c
index 468acd04aa2..47a6f6f1247 100644
--- a/arch/x86/kernel/irq_32.c
+++ b/arch/x86/kernel/irq_32.c
@@ -48,6 +48,29 @@ void ack_bad_irq(unsigned int irq)
 #endif
 }
 
+#ifdef CONFIG_DEBUG_STACKOVERFLOW
+/* Debugging check for stack overflow: is there less than 1KB free? */
+static int check_stack_overflow(void)
+{
+	long sp;
+
+	__asm__ __volatile__("andl %%esp,%0" :
+			     "=r" (sp) : "0" (THREAD_SIZE - 1));
+
+	return sp < (sizeof(struct thread_info) + STACK_WARN);
+}
+
+static void print_stack_overflow(void)
+{
+	printk(KERN_WARNING "low stack detected by irq handler\n");
+	dump_stack();
+}
+
+#else
+static inline int check_stack_overflow(void) { return 0; }
+static inline void print_stack_overflow(void) { }
+#endif
+
 #ifdef CONFIG_4KSTACKS
 /*
  * per-CPU IRQ handling contexts (thread information and stack)
@@ -59,48 +82,29 @@ union irq_ctx {
 
 static union irq_ctx *hardirq_ctx[NR_CPUS] __read_mostly;
 static union irq_ctx *softirq_ctx[NR_CPUS] __read_mostly;
-#endif
 
-/*
- * do_IRQ handles all normal device IRQ's (the special
- * SMP cross-CPU interrupts have their own specific
- * handlers).
- */
-unsigned int do_IRQ(struct pt_regs *regs)
-{	
-	struct pt_regs *old_regs;
-	/* high bit used in ret_from_ code */
-	int irq = ~regs->orig_ax;
-	struct irq_desc *desc = irq_desc + irq;
-#ifdef CONFIG_4KSTACKS
-	union irq_ctx *curctx, *irqctx;
-	u32 *isp;
-#endif
+static char softirq_stack[NR_CPUS * THREAD_SIZE]
+		__attribute__((__section__(".bss.page_aligned")));
 
-	if (unlikely((unsigned)irq >= NR_IRQS)) {
-		printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
-					__func__, irq);
-		BUG();
-	}
+static char hardirq_stack[NR_CPUS * THREAD_SIZE]
+		__attribute__((__section__(".bss.page_aligned")));
 
-	old_regs = set_irq_regs(regs);
-	irq_enter();
-#ifdef CONFIG_DEBUG_STACKOVERFLOW
-	/* Debugging check for stack overflow: is there less than 1KB free? */
-	{
-		long sp;
-
-		__asm__ __volatile__("andl %%esp,%0" :
-					"=r" (sp) : "0" (THREAD_SIZE - 1));
-		if (unlikely(sp < (sizeof(struct thread_info) + STACK_WARN))) {
-			printk("do_IRQ: stack overflow: %ld\n",
-				sp - sizeof(struct thread_info));
-			dump_stack();
-		}
-	}
-#endif
+static void call_on_stack(void *func, void *stack)
+{
+	asm volatile("xchgl	%%ebx,%%esp	\n"
+		     "call	*%%edi		\n"
+		     "movl	%%ebx,%%esp	\n"
+		     : "=b" (stack)
+		     : "0" (stack),
+		       "D"(func)
+		     : "memory", "cc", "edx", "ecx", "eax");
+}
 
-#ifdef CONFIG_4KSTACKS
+static inline int
+execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq)
+{
+	union irq_ctx *curctx, *irqctx;
+	u32 *isp, arg1, arg2;
 
 	curctx = (union irq_ctx *) current_thread_info();
 	irqctx = hardirq_ctx[smp_processor_id()];
@@ -111,52 +115,39 @@ unsigned int do_IRQ(struct pt_regs *regs)
 	 * handler) we can't do that and just have to keep using the
 	 * current stack (which is the irq stack already after all)
 	 */
-	if (curctx != irqctx) {
-		int arg1, arg2, bx;
+	if (unlikely(curctx == irqctx))
+		return 0;
 
-		/* build the stack frame on the IRQ stack */
-		isp = (u32*) ((char*)irqctx + sizeof(*irqctx));
-		irqctx->tinfo.task = curctx->tinfo.task;
-		irqctx->tinfo.previous_esp = current_stack_pointer;
+	/* build the stack frame on the IRQ stack */
+	isp = (u32 *) ((char*)irqctx + sizeof(*irqctx));
+	irqctx->tinfo.task = curctx->tinfo.task;
+	irqctx->tinfo.previous_esp = current_stack_pointer;
 
-		/*
-		 * Copy the softirq bits in preempt_count so that the
-		 * softirq checks work in the hardirq context.
-		 */
-		irqctx->tinfo.preempt_count =
-			(irqctx->tinfo.preempt_count & ~SOFTIRQ_MASK) |
-			(curctx->tinfo.preempt_count & SOFTIRQ_MASK);
-
-		asm volatile(
-			"       xchgl  %%ebx,%%esp    \n"
-			"       call   *%%edi         \n"
-			"       movl   %%ebx,%%esp    \n"
-			: "=a" (arg1), "=d" (arg2), "=b" (bx)
-			:  "0" (irq),   "1" (desc),  "2" (isp),
-			   "D" (desc->handle_irq)
-			: "memory", "cc", "ecx"
-		);
-	} else
-#endif
-		desc->handle_irq(irq, desc);
-
-	irq_exit();
-	set_irq_regs(old_regs);
+	/*
+	 * Copy the softirq bits in preempt_count so that the
+	 * softirq checks work in the hardirq context.
+	 */
+	irqctx->tinfo.preempt_count =
+		(irqctx->tinfo.preempt_count & ~SOFTIRQ_MASK) |
+		(curctx->tinfo.preempt_count & SOFTIRQ_MASK);
+
+	if (unlikely(overflow))
+		call_on_stack(print_stack_overflow, isp);
+
+	asm volatile("xchgl	%%ebx,%%esp	\n"
+		     "call	*%%edi		\n"
+		     "movl	%%ebx,%%esp	\n"
+		     : "=a" (arg1), "=d" (arg2), "=b" (isp)
+		     :  "0" (irq),   "1" (desc),  "2" (isp),
+			"D" (desc->handle_irq)
+		     : "memory", "cc", "ecx");
 	return 1;
 }
 
-#ifdef CONFIG_4KSTACKS
-
-static char softirq_stack[NR_CPUS * THREAD_SIZE]
-		__attribute__((__section__(".bss.page_aligned")));
-
-static char hardirq_stack[NR_CPUS * THREAD_SIZE]
-		__attribute__((__section__(".bss.page_aligned")));
-
 /*
  * allocate per-cpu stacks for hardirq and for softirq processing
  */
-void irq_ctx_init(int cpu)
+void __cpuinit irq_ctx_init(int cpu)
 {
 	union irq_ctx *irqctx;
 
@@ -164,25 +155,25 @@ void irq_ctx_init(int cpu)
 		return;
 
 	irqctx = (union irq_ctx*) &hardirq_stack[cpu*THREAD_SIZE];
-	irqctx->tinfo.task              = NULL;
-	irqctx->tinfo.exec_domain       = NULL;
-	irqctx->tinfo.cpu               = cpu;
-	irqctx->tinfo.preempt_count     = HARDIRQ_OFFSET;
-	irqctx->tinfo.addr_limit        = MAKE_MM_SEG(0);
+	irqctx->tinfo.task		= NULL;
+	irqctx->tinfo.exec_domain	= NULL;
+	irqctx->tinfo.cpu		= cpu;
+	irqctx->tinfo.preempt_count	= HARDIRQ_OFFSET;
+	irqctx->tinfo.addr_limit	= MAKE_MM_SEG(0);
 
 	hardirq_ctx[cpu] = irqctx;
 
 	irqctx = (union irq_ctx*) &softirq_stack[cpu*THREAD_SIZE];
-	irqctx->tinfo.task              = NULL;
-	irqctx->tinfo.exec_domain       = NULL;
-	irqctx->tinfo.cpu               = cpu;
-	irqctx->tinfo.preempt_count     = 0;
-	irqctx->tinfo.addr_limit        = MAKE_MM_SEG(0);
+	irqctx->tinfo.task		= NULL;
+	irqctx->tinfo.exec_domain	= NULL;
+	irqctx->tinfo.cpu		= cpu;
+	irqctx->tinfo.preempt_count	= 0;
+	irqctx->tinfo.addr_limit	= MAKE_MM_SEG(0);
 
 	softirq_ctx[cpu] = irqctx;
 
-	printk("CPU %u irqstacks, hard=%p soft=%p\n",
-		cpu,hardirq_ctx[cpu],softirq_ctx[cpu]);
+	printk(KERN_DEBUG "CPU %u irqstacks, hard=%p soft=%p\n",
+	       cpu,hardirq_ctx[cpu],softirq_ctx[cpu]);
 }
 
 void irq_ctx_exit(int cpu)
@@ -211,25 +202,56 @@ asmlinkage void do_softirq(void)
 		/* build the stack frame on the softirq stack */
 		isp = (u32*) ((char*)irqctx + sizeof(*irqctx));
 
-		asm volatile(
-			"       xchgl   %%ebx,%%esp     \n"
-			"       call    __do_softirq    \n"
-			"       movl    %%ebx,%%esp     \n"
-			: "=b"(isp)
-			: "0"(isp)
-			: "memory", "cc", "edx", "ecx", "eax"
-		);
+		call_on_stack(__do_softirq, isp);
 		/*
 		 * Shouldnt happen, we returned above if in_interrupt():
-	 	 */
+		 */
 		WARN_ON_ONCE(softirq_count());
 	}
 
 	local_irq_restore(flags);
 }
+
+#else
+static inline int
+execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) { return 0; }
 #endif
 
 /*
+ * do_IRQ handles all normal device IRQ's (the special
+ * SMP cross-CPU interrupts have their own specific
+ * handlers).
+ */
+unsigned int do_IRQ(struct pt_regs *regs)
+{
+	struct pt_regs *old_regs;
+	/* high bit used in ret_from_ code */
+	int overflow, irq = ~regs->orig_ax;
+	struct irq_desc *desc = irq_desc + irq;
+
+	if (unlikely((unsigned)irq >= NR_IRQS)) {
+		printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
+					__func__, irq);
+		BUG();
+	}
+
+	old_regs = set_irq_regs(regs);
+	irq_enter();
+
+	overflow = check_stack_overflow();
+
+	if (!execute_on_irq_stack(overflow, desc, irq)) {
+		if (unlikely(overflow))
+			print_stack_overflow();
+		desc->handle_irq(irq, desc);
+	}
+
+	irq_exit();
+	set_irq_regs(old_regs);
+	return 1;
+}
+
+/*
  * Interrupt statistics:
  */
 
diff --git a/arch/x86/kernel/irqinit_64.c b/arch/x86/kernel/irqinit_64.c
index 64bc0f14285..31f49e8f46a 100644
--- a/arch/x86/kernel/irqinit_64.c
+++ b/arch/x86/kernel/irqinit_64.c
@@ -34,6 +34,20 @@
  * interrupt-controller happy.
  */
 
+#define IRQ_NAME2(nr) nr##_interrupt(void)
+#define IRQ_NAME(nr) IRQ_NAME2(IRQ##nr)
+
+/*
+ *	SMP has a few special interrupts for IPI messages
+ */
+
+#define BUILD_IRQ(nr)				\
+	asmlinkage void IRQ_NAME(nr);		\
+	asm("\n.p2align\n"			\
+	    "IRQ" #nr "_interrupt:\n\t"		\
+	    "push $~(" #nr ") ; "		\
+	    "jmp common_interrupt");
+
 #define BI(x,y) \
 	BUILD_IRQ(x##y)
 
@@ -170,33 +184,33 @@ void __init native_init_IRQ(void)
 	 * The reschedule interrupt is a CPU-to-CPU reschedule-helper
 	 * IPI, driven by wakeup.
 	 */
-	set_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt);
+	alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt);
 
 	/* IPIs for invalidation */
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+0, invalidate_interrupt0);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+1, invalidate_interrupt1);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+2, invalidate_interrupt2);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+3, invalidate_interrupt3);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+4, invalidate_interrupt4);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+5, invalidate_interrupt5);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+6, invalidate_interrupt6);
-	set_intr_gate(INVALIDATE_TLB_VECTOR_START+7, invalidate_interrupt7);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+0, invalidate_interrupt0);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+1, invalidate_interrupt1);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+2, invalidate_interrupt2);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+3, invalidate_interrupt3);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+4, invalidate_interrupt4);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+5, invalidate_interrupt5);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+6, invalidate_interrupt6);
+	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+7, invalidate_interrupt7);
 
 	/* IPI for generic function call */
-	set_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt);
+	alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt);
 
 	/* Low priority IPI to cleanup after moving an irq */
 	set_intr_gate(IRQ_MOVE_CLEANUP_VECTOR, irq_move_cleanup_interrupt);
 #endif
-	set_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt);
-	set_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt);
+	alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt);
+	alloc_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt);
 
 	/* self generated IPI for local APIC timer */
-	set_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt);
+	alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt);
 
 	/* IPI vectors for APIC spurious and error interrupts */
-	set_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt);
-	set_intr_gate(ERROR_APIC_VECTOR, error_interrupt);
+	alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt);
+	alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt);
 
 	if (!acpi_ioapic)
 		setup_irq(2, &irq2);
diff --git a/arch/x86/kernel/vmiclock_32.c b/arch/x86/kernel/vmiclock_32.c
index a2b030780aa..ba7d19e102b 100644
--- a/arch/x86/kernel/vmiclock_32.c
+++ b/arch/x86/kernel/vmiclock_32.c
@@ -33,8 +33,7 @@
 #include <asm/apic.h>
 #include <asm/timer.h>
 #include <asm/i8253.h>
-
-#include <irq_vectors.h>
+#include <asm/irq_vectors.h>
 
 #define VMI_ONESHOT  (VMI_ALARM_IS_ONESHOT  | VMI_CYCLES_REAL | vmi_get_alarm_wiring())
 #define VMI_PERIODIC (VMI_ALARM_IS_PERIODIC | VMI_CYCLES_REAL | vmi_get_alarm_wiring())