x86_64: move kernel

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

arch/x86/kernel/Makefile

@@ -1,5 +1,5 @@
 ifeq ($(CONFIG_X86_32),y)
 include ${srctree}/arch/x86/kernel/Makefile_32
 else
-include ${srctree}/arch/x86_64/kernel/Makefile_64
+include ${srctree}/arch/x86/kernel/Makefile_64
 endif

arch/x86/kernel/Makefile_32

@@ -83,6 +83,4 @@ $(obj)/vsyscall-syms.o: $(src)/vsyscall_32.lds \
 			$(obj)/vsyscall-sysenter_32.o $(obj)/vsyscall-note_32.o FORCE
 	$(call if_changed,syscall)
 
-k8-y                      += ../../x86_64/kernel/k8.o
-stacktrace-y		  += ../../x86_64/kernel/stacktrace.o
 

arch/x86/kernel/Makefile_64

@@ -0,0 +1,54 @@
+#
+# Makefile for the linux kernel.
+#
+
+extra-y 	:= head_64.o head64.o init_task_64.o vmlinux.lds
+EXTRA_AFLAGS	:= -traditional
+obj-y	:= process_64.o signal_64.o entry_64.o traps_64.o irq_64.o \
+		ptrace_64.o time_64.o ioport_64.o ldt_64.o setup_64.o i8259_64.o sys_x86_64.o \
+		x8664_ksyms_64.o i387_64.o syscall_64.o vsyscall_64.o \
+		setup64.o bootflag.o e820_64.o reboot_64.o quirks.o i8237.o \
+		pci-dma_64.o pci-nommu_64.o alternative.o hpet_64.o tsc_64.o bugs_64.o \
+		perfctr-watchdog.o
+
+obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
+obj-$(CONFIG_X86_MCE)		+= mce_64.o therm_throt.o
+obj-$(CONFIG_X86_MCE_INTEL)	+= mce_intel_64.o
+obj-$(CONFIG_X86_MCE_AMD)	+= mce_amd_64.o
+obj-$(CONFIG_MTRR)		+= ../../x86/kernel/cpu/mtrr/
+obj-$(CONFIG_ACPI)		+= ../../x86/kernel/acpi/
+obj-$(CONFIG_X86_MSR)		+= msr.o
+obj-$(CONFIG_MICROCODE)		+= microcode.o
+obj-$(CONFIG_X86_CPUID)		+= cpuid.o
+obj-$(CONFIG_SMP)		+= smp_64.o smpboot_64.o trampoline_64.o tsc_sync.o
+obj-y				+= apic_64.o  nmi_64.o
+obj-y				+= io_apic_64.o mpparse_64.o genapic_64.o genapic_flat_64.o
+obj-$(CONFIG_KEXEC)		+= machine_kexec_64.o relocate_kernel_64.o crash_64.o
+obj-$(CONFIG_CRASH_DUMP)	+= crash_dump_64.o
+obj-$(CONFIG_PM)		+= suspend_64.o
+obj-$(CONFIG_HIBERNATION)	+= suspend_asm_64.o
+obj-$(CONFIG_CPU_FREQ)		+= ../../x86/kernel/cpu/cpufreq/
+obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
+obj-$(CONFIG_IOMMU)		+= pci-gart_64.o aperture_64.o
+obj-$(CONFIG_CALGARY_IOMMU)	+= pci-calgary_64.o tce_64.o
+obj-$(CONFIG_SWIOTLB)		+= pci-swiotlb_64.o
+obj-$(CONFIG_KPROBES)		+= kprobes_64.o
+obj-$(CONFIG_X86_PM_TIMER)	+= pmtimer_64.o
+obj-$(CONFIG_X86_VSMP)		+= vsmp_64.o
+obj-$(CONFIG_K8_NB)		+= k8.o
+obj-$(CONFIG_AUDIT)		+= audit_64.o
+
+obj-$(CONFIG_MODULES)		+= module_64.o
+obj-$(CONFIG_PCI)		+= early-quirks_64.o
+
+obj-y				+= topology.o
+obj-y				+= intel_cacheinfo.o
+obj-y				+= addon_cpuid_features.o
+obj-y				+= pcspeaker.o
+
+CFLAGS_vsyscall_64.o		:= $(PROFILING) -g0
+
+therm_throt-y                   += ../../x86/kernel/cpu/mcheck/therm_throt.o
+intel_cacheinfo-y		+= ../../x86/kernel/cpu/intel_cacheinfo.o
+addon_cpuid_features-y		+= ../../x86/kernel/cpu/addon_cpuid_features.o
+perfctr-watchdog-y		+= ../../x86/kernel/cpu/perfctr-watchdog.o

arch/x86/kernel/acpi/wakeup_64.S

@@ -269,7 +269,7 @@ no_longmode:
 	movb    $0xbc,%al       ;  outb %al,$0x80
 	jmp no_longmode
 
-#include "../../../x86_64/kernel/verify_cpu_64.S"
+#include "../verify_cpu_64.S"
 	
 /* This code uses an extended set of video mode numbers. These include:
  * Aliases for standard modes

arch/x86/kernel/aperture_64.c

@@ -0,0 +1,298 @@
+/* 
+ * Firmware replacement code.
+ * 
+ * Work around broken BIOSes that don't set an aperture or only set the
+ * aperture in the AGP bridge. 
+ * If all fails map the aperture over some low memory.  This is cheaper than 
+ * doing bounce buffering. The memory is lost. This is done at early boot 
+ * because only the bootmem allocator can allocate 32+MB. 
+ * 
+ * Copyright 2002 Andi Kleen, SuSE Labs.
+ */
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/bootmem.h>
+#include <linux/mmzone.h>
+#include <linux/pci_ids.h>
+#include <linux/pci.h>
+#include <linux/bitops.h>
+#include <linux/ioport.h>
+#include <asm/e820.h>
+#include <asm/io.h>
+#include <asm/iommu.h>
+#include <asm/pci-direct.h>
+#include <asm/dma.h>
+#include <asm/k8.h>
+
+int iommu_aperture;
+int iommu_aperture_disabled __initdata = 0;
+int iommu_aperture_allowed __initdata = 0;
+
+int fallback_aper_order __initdata = 1; /* 64MB */
+int fallback_aper_force __initdata = 0; 
+
+int fix_aperture __initdata = 1;
+
+static struct resource gart_resource = {
+	.name	= "GART",
+	.flags	= IORESOURCE_MEM,
+};
+
+static void __init insert_aperture_resource(u32 aper_base, u32 aper_size)
+{
+	gart_resource.start = aper_base;
+	gart_resource.end = aper_base + aper_size - 1;
+	insert_resource(&iomem_resource, &gart_resource);
+}
+
+/* This code runs before the PCI subsystem is initialized, so just
+   access the northbridge directly. */
+
+static u32 __init allocate_aperture(void) 
+{
+	u32 aper_size;
+	void *p; 
+
+	if (fallback_aper_order > 7) 
+		fallback_aper_order = 7; 
+	aper_size = (32 * 1024 * 1024) << fallback_aper_order; 
+
+	/* 
+	 * Aperture has to be naturally aligned. This means an 2GB aperture won't
+	 * have much chance of finding a place in the lower 4GB of memory.
+	 * Unfortunately we cannot move it up because that would make the
+	 * IOMMU useless.
+	 */
+	p = __alloc_bootmem_nopanic(aper_size, aper_size, 0);
+	if (!p || __pa(p)+aper_size > 0xffffffff) {
+		printk("Cannot allocate aperture memory hole (%p,%uK)\n",
+		       p, aper_size>>10);
+		if (p)
+			free_bootmem(__pa(p), aper_size);
+		return 0;
+	}
+	printk("Mapping aperture over %d KB of RAM @ %lx\n",
+	       aper_size >> 10, __pa(p)); 
+	insert_aperture_resource((u32)__pa(p), aper_size);
+	return (u32)__pa(p); 
+}
+
+static int __init aperture_valid(u64 aper_base, u32 aper_size)
+{ 
+	if (!aper_base) 
+		return 0;
+	if (aper_size < 64*1024*1024) { 
+		printk("Aperture too small (%d MB)\n", aper_size>>20);
+		return 0;
+	}
+	if (aper_base + aper_size > 0x100000000UL) {
+		printk("Aperture beyond 4GB. Ignoring.\n");
+		return 0; 
+	}
+	if (e820_any_mapped(aper_base, aper_base + aper_size, E820_RAM)) {
+		printk("Aperture pointing to e820 RAM. Ignoring.\n");
+		return 0; 
+	} 
+	return 1;
+} 
+
+/* Find a PCI capability */
+static __u32 __init find_cap(int num, int slot, int func, int cap) 
+{ 
+	u8 pos;
+	int bytes;
+	if (!(read_pci_config_16(num,slot,func,PCI_STATUS) & PCI_STATUS_CAP_LIST))
+		return 0;
+	pos = read_pci_config_byte(num,slot,func,PCI_CAPABILITY_LIST);
+	for (bytes = 0; bytes < 48 && pos >= 0x40; bytes++) { 
+		u8 id;
+		pos &= ~3; 
+		id = read_pci_config_byte(num,slot,func,pos+PCI_CAP_LIST_ID);
+		if (id == 0xff)
+			break;
+		if (id == cap) 
+			return pos; 
+		pos = read_pci_config_byte(num,slot,func,pos+PCI_CAP_LIST_NEXT); 
+	} 
+	return 0;
+} 
+
+/* Read a standard AGPv3 bridge header */
+static __u32 __init read_agp(int num, int slot, int func, int cap, u32 *order)
+{ 
+	u32 apsize;
+	u32 apsizereg;
+	int nbits;
+	u32 aper_low, aper_hi;
+	u64 aper;
+
+	printk("AGP bridge at %02x:%02x:%02x\n", num, slot, func);
+	apsizereg = read_pci_config_16(num,slot,func, cap + 0x14);
+	if (apsizereg == 0xffffffff) {
+		printk("APSIZE in AGP bridge unreadable\n");
+		return 0;
+	}
+
+	apsize = apsizereg & 0xfff;
+	/* Some BIOS use weird encodings not in the AGPv3 table. */
+	if (apsize & 0xff) 
+		apsize |= 0xf00; 
+	nbits = hweight16(apsize);
+	*order = 7 - nbits;
+	if ((int)*order < 0) /* < 32MB */
+		*order = 0;
+	
+	aper_low = read_pci_config(num,slot,func, 0x10);
+	aper_hi = read_pci_config(num,slot,func,0x14);
+	aper = (aper_low & ~((1<<22)-1)) | ((u64)aper_hi << 32);
+
+	printk("Aperture from AGP @ %Lx size %u MB (APSIZE %x)\n", 
+	       aper, 32 << *order, apsizereg);
+
+	if (!aperture_valid(aper, (32*1024*1024) << *order))
+	    return 0;
+	return (u32)aper; 
+} 
+
+/* Look for an AGP bridge. Windows only expects the aperture in the
+   AGP bridge and some BIOS forget to initialize the Northbridge too.
+   Work around this here. 
+
+   Do an PCI bus scan by hand because we're running before the PCI
+   subsystem. 
+
+   All K8 AGP bridges are AGPv3 compliant, so we can do this scan
+   generically. It's probably overkill to always scan all slots because
+   the AGP bridges should be always an own bus on the HT hierarchy, 
+   but do it here for future safety. */
+static __u32 __init search_agp_bridge(u32 *order, int *valid_agp)
+{
+	int num, slot, func;
+
+	/* Poor man's PCI discovery */
+	for (num = 0; num < 256; num++) { 
+		for (slot = 0; slot < 32; slot++) { 
+			for (func = 0; func < 8; func++) { 
+				u32 class, cap;
+				u8 type;
+				class = read_pci_config(num,slot,func,
+							PCI_CLASS_REVISION);
+				if (class == 0xffffffff)
+					break; 
+				
+				switch (class >> 16) { 
+				case PCI_CLASS_BRIDGE_HOST:
+				case PCI_CLASS_BRIDGE_OTHER: /* needed? */
+					/* AGP bridge? */
+					cap = find_cap(num,slot,func,PCI_CAP_ID_AGP);
+					if (!cap)
+						break;
+					*valid_agp = 1; 
+					return read_agp(num,slot,func,cap,order);
+				} 
+				
+				/* No multi-function device? */
+				type = read_pci_config_byte(num,slot,func,
+							       PCI_HEADER_TYPE);
+				if (!(type & 0x80))
+					break;
+			} 
+		} 
+	}
+	printk("No AGP bridge found\n"); 
+	return 0;
+}
+
+void __init iommu_hole_init(void) 
+{ 
+	int fix, num; 
+	u32 aper_size, aper_alloc = 0, aper_order = 0, last_aper_order = 0;
+	u64 aper_base, last_aper_base = 0;
+	int valid_agp = 0;
+
+	if (iommu_aperture_disabled || !fix_aperture || !early_pci_allowed())
+		return;
+
+	printk(KERN_INFO  "Checking aperture...\n");
+
+	fix = 0;
+	for (num = 24; num < 32; num++) {		
+		if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00)))
+			continue;
+
+		iommu_detected = 1;
+		iommu_aperture = 1; 
+
+		aper_order = (read_pci_config(0, num, 3, 0x90) >> 1) & 7; 
+		aper_size = (32 * 1024 * 1024) << aper_order; 
+		aper_base = read_pci_config(0, num, 3, 0x94) & 0x7fff;
+		aper_base <<= 25; 
+
+		printk("CPU %d: aperture @ %Lx size %u MB\n", num-24, 
+		       aper_base, aper_size>>20);
+		
+		if (!aperture_valid(aper_base, aper_size)) {
+			fix = 1; 
+			break; 
+		}
+
+		if ((last_aper_order && aper_order != last_aper_order) ||
+		    (last_aper_base && aper_base != last_aper_base)) {
+			fix = 1;
+			break;
+		}
+		last_aper_order = aper_order;
+		last_aper_base = aper_base;
+	} 
+
+	if (!fix && !fallback_aper_force) {
+		if (last_aper_base) {
+			unsigned long n = (32 * 1024 * 1024) << last_aper_order;
+			insert_aperture_resource((u32)last_aper_base, n);
+		}
+		return; 
+	}
+
+	if (!fallback_aper_force)
+		aper_alloc = search_agp_bridge(&aper_order, &valid_agp); 
+		
+	if (aper_alloc) { 
+		/* Got the aperture from the AGP bridge */
+	} else if (swiotlb && !valid_agp) {
+		/* Do nothing */
+	} else if ((!no_iommu && end_pfn > MAX_DMA32_PFN) ||
+		   force_iommu ||
+		   valid_agp ||
+		   fallback_aper_force) { 
+		printk("Your BIOS doesn't leave a aperture memory hole\n");
+		printk("Please enable the IOMMU option in the BIOS setup\n");
+		printk("This costs you %d MB of RAM\n",
+		       32 << fallback_aper_order);
+
+		aper_order = fallback_aper_order;
+		aper_alloc = allocate_aperture();
+		if (!aper_alloc) { 
+			/* Could disable AGP and IOMMU here, but it's probably
+			   not worth it. But the later users cannot deal with
+			   bad apertures and turning on the aperture over memory
+			   causes very strange problems, so it's better to 
+			   panic early. */
+			panic("Not enough memory for aperture");
+		}
+	} else { 
+		return; 
+	} 
+
+	/* Fix up the north bridges */
+	for (num = 24; num < 32; num++) { 		
+		if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00)))
+			continue;	
+
+		/* Don't enable translation yet. That is done later. 
+		   Assume this BIOS didn't initialise the GART so 
+		   just overwrite all previous bits */ 
+		write_pci_config(0, num, 3, 0x90, aper_order<<1); 
+		write_pci_config(0, num, 3, 0x94, aper_alloc>>25); 
+	} 
+} 

arch/x86/kernel/asm-offsets_64.c

@@ -0,0 +1,85 @@
+/*
+ * Generate definitions needed by assembly language modules.
+ * This code generates raw asm output which is post-processed to extract
+ * and format the required data.
+ */
+
+#include <linux/crypto.h>
+#include <linux/sched.h> 
+#include <linux/stddef.h>
+#include <linux/errno.h> 
+#include <linux/hardirq.h>
+#include <linux/suspend.h>
+#include <asm/pda.h>
+#include <asm/processor.h>
+#include <asm/segment.h>
+#include <asm/thread_info.h>
+#include <asm/ia32.h>
+
+#define DEFINE(sym, val) \
+        asm volatile("\n->" #sym " %0 " #val : : "i" (val))
+
+#define BLANK() asm volatile("\n->" : : )
+
+#define __NO_STUBS 1
+#undef __SYSCALL
+#undef _ASM_X86_64_UNISTD_H_
+#define __SYSCALL(nr, sym) [nr] = 1,
+static char syscalls[] = {
+#include <asm/unistd.h>
+};
+
+int main(void)
+{
+#define ENTRY(entry) DEFINE(tsk_ ## entry, offsetof(struct task_struct, entry))
+	ENTRY(state);
+	ENTRY(flags); 
+	ENTRY(thread); 
+	ENTRY(pid);
+	BLANK();
+#undef ENTRY
+#define ENTRY(entry) DEFINE(threadinfo_ ## entry, offsetof(struct thread_info, entry))
+	ENTRY(flags);
+	ENTRY(addr_limit);
+	ENTRY(preempt_count);
+	ENTRY(status);
+	BLANK();
+#undef ENTRY
+#define ENTRY(entry) DEFINE(pda_ ## entry, offsetof(struct x8664_pda, entry))
+	ENTRY(kernelstack); 
+	ENTRY(oldrsp); 
+	ENTRY(pcurrent); 
+	ENTRY(irqcount);
+	ENTRY(cpunumber);
+	ENTRY(irqstackptr);
+	ENTRY(data_offset);
+	BLANK();
+#undef ENTRY
+#ifdef CONFIG_IA32_EMULATION
+#define ENTRY(entry) DEFINE(IA32_SIGCONTEXT_ ## entry, offsetof(struct sigcontext_ia32, entry))
+	ENTRY(eax);
+	ENTRY(ebx);
+	ENTRY(ecx);
+	ENTRY(edx);
+	ENTRY(esi);
+	ENTRY(edi);
+	ENTRY(ebp);
+	ENTRY(esp);
+	ENTRY(eip);
+	BLANK();
+#undef ENTRY
+	DEFINE(IA32_RT_SIGFRAME_sigcontext,
+	       offsetof (struct rt_sigframe32, uc.uc_mcontext));
+	BLANK();
+#endif
+	DEFINE(pbe_address, offsetof(struct pbe, address));
+	DEFINE(pbe_orig_address, offsetof(struct pbe, orig_address));
+	DEFINE(pbe_next, offsetof(struct pbe, next));
+	BLANK();
+	DEFINE(TSS_ist, offsetof(struct tss_struct, ist));
+	BLANK();
+	DEFINE(crypto_tfm_ctx_offset, offsetof(struct crypto_tfm, __crt_ctx));
+	BLANK();
+	DEFINE(__NR_syscall_max, sizeof(syscalls) - 1);
+	return 0;
+}

arch/x86/kernel/audit_64.c

@@ -0,0 +1,81 @@
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/audit.h>
+#include <asm/unistd.h>
+
+static unsigned dir_class[] = {
+#include <asm-generic/audit_dir_write.h>
+~0U
+};
+
+static unsigned read_class[] = {
+#include <asm-generic/audit_read.h>
+~0U
+};
+
+static unsigned write_class[] = {
+#include <asm-generic/audit_write.h>
+~0U
+};
+
+static unsigned chattr_class[] = {
+#include <asm-generic/audit_change_attr.h>
+~0U
+};
+
+static unsigned signal_class[] = {
+#include <asm-generic/audit_signal.h>
+~0U
+};
+
+int audit_classify_arch(int arch)
+{
+#ifdef CONFIG_IA32_EMULATION
+	if (arch == AUDIT_ARCH_I386)
+		return 1;
+#endif
+	return 0;
+}
+
+int audit_classify_syscall(int abi, unsigned syscall)
+{
+#ifdef CONFIG_IA32_EMULATION
+	extern int ia32_classify_syscall(unsigned);
+	if (abi == AUDIT_ARCH_I386)
+		return ia32_classify_syscall(syscall);
+#endif
+	switch(syscall) {
+	case __NR_open:
+		return 2;
+	case __NR_openat:
+		return 3;
+	case __NR_execve:
+		return 5;
+	default:
+		return 0;
+	}
+}
+
+static int __init audit_classes_init(void)
+{
+#ifdef CONFIG_IA32_EMULATION
+	extern __u32 ia32_dir_class[];
+	extern __u32 ia32_write_class[];
+	extern __u32 ia32_read_class[];
+	extern __u32 ia32_chattr_class[];
+	extern __u32 ia32_signal_class[];
+	audit_register_class(AUDIT_CLASS_WRITE_32, ia32_write_class);
+	audit_register_class(AUDIT_CLASS_READ_32, ia32_read_class);
+	audit_register_class(AUDIT_CLASS_DIR_WRITE_32, ia32_dir_class);
+	audit_register_class(AUDIT_CLASS_CHATTR_32, ia32_chattr_class);
+	audit_register_class(AUDIT_CLASS_SIGNAL_32, ia32_signal_class);
+#endif
+	audit_register_class(AUDIT_CLASS_WRITE, write_class);
+	audit_register_class(AUDIT_CLASS_READ, read_class);
+	audit_register_class(AUDIT_CLASS_DIR_WRITE, dir_class);
+	audit_register_class(AUDIT_CLASS_CHATTR, chattr_class);
+	audit_register_class(AUDIT_CLASS_SIGNAL, signal_class);
+	return 0;
+}
+
+__initcall(audit_classes_init);

arch/x86/kernel/bugs_64.c

@@ -0,0 +1,24 @@
+/*
+ *  arch/x86_64/kernel/bugs.c
+ *
+ *  Copyright (C) 1994  Linus Torvalds
+ *  Copyright (C) 2000  SuSE
+ */
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <asm/alternative.h>
+#include <asm/bugs.h>
+#include <asm/processor.h>
+#include <asm/mtrr.h>
+
+void __init check_bugs(void)
+{
+	identify_cpu(&boot_cpu_data);
+	mtrr_bp_init();
+#if !defined(CONFIG_SMP)
+	printk("CPU: ");
+	print_cpu_info(&boot_cpu_data);
+#endif
+	alternative_instructions();
+}

arch/x86/kernel/crash_64.c

@@ -0,0 +1,135 @@
+/*
+ * Architecture specific (x86_64) functions for kexec based crash dumps.
+ *
+ * Created by: Hariprasad Nellitheertha (hari@in.ibm.com)
+ *
+ * Copyright (C) IBM Corporation, 2004. All rights reserved.
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/smp.h>
+#include <linux/irq.h>
+#include <linux/reboot.h>
+#include <linux/kexec.h>
+#include <linux/delay.h>
+#include <linux/elf.h>
+#include <linux/elfcore.h>
+#include <linux/kdebug.h>
+
+#include <asm/processor.h>
+#include <asm/hardirq.h>
+#include <asm/nmi.h>
+#include <asm/hw_irq.h>
+#include <asm/mach_apic.h>
+
+/* This keeps a track of which one is crashing cpu. */
+static int crashing_cpu;
+
+#ifdef CONFIG_SMP
+static atomic_t waiting_for_crash_ipi;
+
+static int crash_nmi_callback(struct notifier_block *self,
+				unsigned long val, void *data)
+{
+	struct pt_regs *regs;
+	int cpu;
+
+	if (val != DIE_NMI_IPI)
+		return NOTIFY_OK;
+
+	regs = ((struct die_args *)data)->regs;
+	cpu = raw_smp_processor_id();
+
+	/*
+	 * Don't do anything if this handler is invoked on crashing cpu.
+	 * Otherwise, system will completely hang. Crashing cpu can get
+	 * an NMI if system was initially booted with nmi_watchdog parameter.
+	 */
+	if (cpu == crashing_cpu)
+		return NOTIFY_STOP;
+	local_irq_disable();
+
+	crash_save_cpu(regs, cpu);
+	disable_local_APIC();
+	atomic_dec(&waiting_for_crash_ipi);
+	/* Assume hlt works */
+	for(;;)
+		halt();
+
+	return 1;
+}
+
+static void smp_send_nmi_allbutself(void)
+{
+	send_IPI_allbutself(NMI_VECTOR);
+}
+
+/*
+ * This code is a best effort heuristic to get the
+ * other cpus to stop executing. So races with
+ * cpu hotplug shouldn't matter.
+ */
+
+static struct notifier_block crash_nmi_nb = {
+	.notifier_call = crash_nmi_callback,
+};
+
+static void nmi_shootdown_cpus(void)
+{
+	unsigned long msecs;
+
+	atomic_set(&waiting_for_crash_ipi, num_online_cpus() - 1);
+	if (register_die_notifier(&crash_nmi_nb))
+		return;         /* return what? */
+
+	/*
+	 * Ensure the new callback function is set before sending
+	 * out the NMI
+	 */
+	wmb();
+
+	smp_send_nmi_allbutself();
+
+	msecs = 1000; /* Wait at most a second for the other cpus to stop */
+	while ((atomic_read(&waiting_for_crash_ipi) > 0) && msecs) {
+		mdelay(1);
+		msecs--;
+	}
+	/* Leave the nmi callback set */
+	disable_local_APIC();
+}
+#else
+static void nmi_shootdown_cpus(void)
+{
+	/* There are no cpus to shootdown */
+}
+#endif
+
+void machine_crash_shutdown(struct pt_regs *regs)
+{
+	/*
+	 * This function is only called after the system
+	 * has panicked or is otherwise in a critical state.
+	 * The minimum amount of code to allow a kexec'd kernel
+	 * to run successfully needs to happen here.
+	 *
+	 * In practice this means shooting down the other cpus in
+	 * an SMP system.
+	 */
+	/* The kernel is broken so disable interrupts */
+	local_irq_disable();
+
+	/* Make a note of crashing cpu. Will be used in NMI callback.*/
+	crashing_cpu = smp_processor_id();
+	nmi_shootdown_cpus();
+
+	if(cpu_has_apic)
+		 disable_local_APIC();
+
+	disable_IO_APIC();
+
+	crash_save_cpu(regs, smp_processor_id());
+}

arch/x86/kernel/crash_dump_64.c

@@ -0,0 +1,47 @@
+/*
+ *	kernel/crash_dump.c - Memory preserving reboot related code.
+ *
+ *	Created by: Hariprasad Nellitheertha (hari@in.ibm.com)
+ *	Copyright (C) IBM Corporation, 2004. All rights reserved
+ */
+
+#include <linux/errno.h>
+#include <linux/crash_dump.h>
+
+#include <asm/uaccess.h>
+#include <asm/io.h>
+
+/**
+ * copy_oldmem_page - copy one page from "oldmem"
+ * @pfn: page frame number to be copied
+ * @buf: target memory address for the copy; this can be in kernel address
+ *	space or user address space (see @userbuf)
+ * @csize: number of bytes to copy
+ * @offset: offset in bytes into the page (based on pfn) to begin the copy
+ * @userbuf: if set, @buf is in user address space, use copy_to_user(),
+ *	otherwise @buf is in kernel address space, use memcpy().
+ *
+ * Copy a page from "oldmem". For this page, there is no pte mapped
+ * in the current kernel. We stitch up a pte, similar to kmap_atomic.
+ */
+ssize_t copy_oldmem_page(unsigned long pfn, char *buf,
+                               size_t csize, unsigned long offset, int userbuf)
+{
+	void  *vaddr;
+
+	if (!csize)
+		return 0;
+
+	vaddr = ioremap(pfn << PAGE_SHIFT, PAGE_SIZE);
+
+	if (userbuf) {
+		if (copy_to_user(buf, (vaddr + offset), csize)) {
+			iounmap(vaddr);
+			return -EFAULT;
+		}
+	} else
+	memcpy(buf, (vaddr + offset), csize);
+
+	iounmap(vaddr);
+	return csize;
+}

arch/x86/kernel/e820_64.c

@@ -0,0 +1,725 @@
+/* 
+ * Handle the memory map.
+ * The functions here do the job until bootmem takes over.
+ *
+ *  Getting sanitize_e820_map() in sync with i386 version by applying change:
+ *  -  Provisions for empty E820 memory regions (reported by certain BIOSes).
+ *     Alex Achenbach <xela@slit.de>, December 2002.
+ *  Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
+ *
+ */
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/bootmem.h>
+#include <linux/ioport.h>
+#include <linux/string.h>
+#include <linux/kexec.h>
+#include <linux/module.h>
+#include <linux/mm.h>
+#include <linux/suspend.h>
+#include <linux/pfn.h>
+
+#include <asm/pgtable.h>
+#include <asm/page.h>
+#include <asm/e820.h>
+#include <asm/proto.h>
+#include <asm/bootsetup.h>
+#include <asm/sections.h>
+
+struct e820map e820;
+
+/* 
+ * PFN of last memory page.
+ */
+unsigned long end_pfn; 
+EXPORT_SYMBOL(end_pfn);
+
+/* 
+ * end_pfn only includes RAM, while end_pfn_map includes all e820 entries.
+ * The direct mapping extends to end_pfn_map, so that we can directly access
+ * apertures, ACPI and other tables without having to play with fixmaps.
+ */ 
+unsigned long end_pfn_map; 
+
+/* 
+ * Last pfn which the user wants to use.
+ */
+static unsigned long __initdata end_user_pfn = MAXMEM>>PAGE_SHIFT;
+
+extern struct resource code_resource, data_resource;
+
+/* Check for some hardcoded bad areas that early boot is not allowed to touch */ 
+static inline int bad_addr(unsigned long *addrp, unsigned long size)
+{ 
+	unsigned long addr = *addrp, last = addr + size; 
+
+	/* various gunk below that needed for SMP startup */
+	if (addr < 0x8000) { 
+		*addrp = PAGE_ALIGN(0x8000);
+		return 1; 
+	}
+
+	/* direct mapping tables of the kernel */
+	if (last >= table_start<<PAGE_SHIFT && addr < table_end<<PAGE_SHIFT) { 
+		*addrp = PAGE_ALIGN(table_end << PAGE_SHIFT);
+		return 1;
+	} 
+
+	/* initrd */ 
+#ifdef CONFIG_BLK_DEV_INITRD
+	if (LOADER_TYPE && INITRD_START && last >= INITRD_START && 
+	    addr < INITRD_START+INITRD_SIZE) { 
+		*addrp = PAGE_ALIGN(INITRD_START + INITRD_SIZE);
+		return 1;
+	} 
+#endif
+	/* kernel code */
+	if (last >= __pa_symbol(&_text) && addr < __pa_symbol(&_end)) {
+		*addrp = PAGE_ALIGN(__pa_symbol(&_end));
+		return 1;
+	}
+
+	if (last >= ebda_addr && addr < ebda_addr + ebda_size) {
+		*addrp = PAGE_ALIGN(ebda_addr + ebda_size);
+		return 1;
+	}
+
+#ifdef CONFIG_NUMA
+	/* NUMA memory to node map */
+	if (last >= nodemap_addr && addr < nodemap_addr + nodemap_size) {
+		*addrp = nodemap_addr + nodemap_size;
+		return 1;
+	}
+#endif
+	/* XXX ramdisk image here? */ 
+	return 0;
+} 
+
+/*
+ * This function checks if any part of the range <start,end> is mapped
+ * with type.
+ */
+int
+e820_any_mapped(unsigned long start, unsigned long end, unsigned type)
+{ 
+	int i;
+	for (i = 0; i < e820.nr_map; i++) { 
+		struct e820entry *ei = &e820.map[i]; 
+		if (type && ei->type != type) 
+			continue;
+		if (ei->addr >= end || ei->addr + ei->size <= start)
+			continue; 
+		return 1; 
+	} 
+	return 0;
+}
+EXPORT_SYMBOL_GPL(e820_any_mapped);
+
+/*
+ * This function checks if the entire range <start,end> is mapped with type.
+ *
+ * Note: this function only works correct if the e820 table is sorted and
+ * not-overlapping, which is the case
+ */
+int __init e820_all_mapped(unsigned long start, unsigned long end, unsigned type)
+{
+	int i;
+	for (i = 0; i < e820.nr_map; i++) {
+		struct e820entry *ei = &e820.map[i];
+		if (type && ei->type != type)
+			continue;
+		/* is the region (part) in overlap with the current region ?*/
+		if (ei->addr >= end || ei->addr + ei->size <= start)
+			continue;
+
+		/* if the region is at the beginning of <start,end> we move
+		 * start to the end of the region since it's ok until there
+		 */
+		if (ei->addr <= start)
+			start = ei->addr + ei->size;
+		/* if start is now at or beyond end, we're done, full coverage */
+		if (start >= end)
+			return 1; /* we're done */
+	}
+	return 0;
+}
+
+/* 
+ * Find a free area in a specific range. 
+ */ 
+unsigned long __init find_e820_area(unsigned long start, unsigned long end, unsigned size) 
+{ 
+	int i; 
+	for (i = 0; i < e820.nr_map; i++) { 
+		struct e820entry *ei = &e820.map[i]; 
+		unsigned long addr = ei->addr, last; 
+		if (ei->type != E820_RAM) 
+			continue; 
+		if (addr < start) 
+			addr = start;
+		if (addr > ei->addr + ei->size) 
+			continue; 
+		while (bad_addr(&addr, size) && addr+size <= ei->addr+ei->size)
+			;
+		last = PAGE_ALIGN(addr) + size;
+		if (last > ei->addr + ei->size)
+			continue;
+		if (last > end) 
+			continue;
+		return addr; 
+	} 
+	return -1UL;		
+} 
+
+/*
+ * Find the highest page frame number we have available
+ */
+unsigned long __init e820_end_of_ram(void)
+{
+	unsigned long end_pfn = 0;
+	end_pfn = find_max_pfn_with_active_regions();
+	
+	if (end_pfn > end_pfn_map) 
+		end_pfn_map = end_pfn;
+	if (end_pfn_map > MAXMEM>>PAGE_SHIFT)
+		end_pfn_map = MAXMEM>>PAGE_SHIFT;
+	if (end_pfn > end_user_pfn)
+		end_pfn = end_user_pfn;
+	if (end_pfn > end_pfn_map) 
+		end_pfn = end_pfn_map; 
+
+	printk("end_pfn_map = %lu\n", end_pfn_map);
+	return end_pfn;	
+}
+
+/*
+ * Mark e820 reserved areas as busy for the resource manager.
+ */
+void __init e820_reserve_resources(void)
+{
+	int i;
+	for (i = 0; i < e820.nr_map; i++) {
+		struct resource *res;
+		res = alloc_bootmem_low(sizeof(struct resource));
+		switch (e820.map[i].type) {
+		case E820_RAM:	res->name = "System RAM"; break;
+		case E820_ACPI:	res->name = "ACPI Tables"; break;
+		case E820_NVS:	res->name = "ACPI Non-volatile Storage"; break;
+		default:	res->name = "reserved";
+		}
+		res->start = e820.map[i].addr;
+		res->end = res->start + e820.map[i].size - 1;
+		res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
+		request_resource(&iomem_resource, res);
+		if (e820.map[i].type == E820_RAM) {
+			/*
+			 *  We don't know which RAM region contains kernel data,
+			 *  so we try it repeatedly and let the resource manager
+			 *  test it.
+			 */
+			request_resource(res, &code_resource);
+			request_resource(res, &data_resource);
+#ifdef CONFIG_KEXEC
+			request_resource(res, &crashk_res);
+#endif
+		}
+	}
+}
+
+/*
+ * Find the ranges of physical addresses that do not correspond to
+ * e820 RAM areas and mark the corresponding pages as nosave for software
+ * suspend and suspend to RAM.
+ *
+ * This function requires the e820 map to be sorted and without any
+ * overlapping entries and assumes the first e820 area to be RAM.
+ */
+void __init e820_mark_nosave_regions(void)
+{
+	int i;
+	unsigned long paddr;
+
+	paddr = round_down(e820.map[0].addr + e820.map[0].size, PAGE_SIZE);
+	for (i = 1; i < e820.nr_map; i++) {
+		struct e820entry *ei = &e820.map[i];
+
+		if (paddr < ei->addr)
+			register_nosave_region(PFN_DOWN(paddr),
+						PFN_UP(ei->addr));
+
+		paddr = round_down(ei->addr + ei->size, PAGE_SIZE);
+		if (ei->type != E820_RAM)
+			register_nosave_region(PFN_UP(ei->addr),
+						PFN_DOWN(paddr));
+
+		if (paddr >= (end_pfn << PAGE_SHIFT))
+			break;
+	}
+}
+
+/*
+ * Finds an active region in the address range from start_pfn to end_pfn and
+ * returns its range in ei_startpfn and ei_endpfn for the e820 entry.
+ */
+static int __init e820_find_active_region(const struct e820entry *ei,
+					  unsigned long start_pfn,
+					  unsigned long end_pfn,
+					  unsigned long *ei_startpfn,
+					  unsigned long *ei_endpfn)
+{
+	*ei_startpfn = round_up(ei->addr, PAGE_SIZE) >> PAGE_SHIFT;
+	*ei_endpfn = round_down(ei->addr + ei->size, PAGE_SIZE) >> PAGE_SHIFT;
+
+	/* Skip map entries smaller than a page */
+	if (*ei_startpfn >= *ei_endpfn)
+		return 0;
+
+	/* Check if end_pfn_map should be updated */
+	if (ei->type != E820_RAM && *ei_endpfn > end_pfn_map)
+		end_pfn_map = *ei_endpfn;
+
+	/* Skip if map is outside the node */
+	if (ei->type != E820_RAM || *ei_endpfn <= start_pfn ||
+				    *ei_startpfn >= end_pfn)
+		return 0;
+
+	/* Check for overlaps */
+	if (*ei_startpfn < start_pfn)
+		*ei_startpfn = start_pfn;
+	if (*ei_endpfn > end_pfn)
+		*ei_endpfn = end_pfn;
+
+	/* Obey end_user_pfn to save on memmap */
+	if (*ei_startpfn >= end_user_pfn)
+		return 0;
+	if (*ei_endpfn > end_user_pfn)
+		*ei_endpfn = end_user_pfn;
+
+	return 1;
+}
+
+/* Walk the e820 map and register active regions within a node */
+void __init
+e820_register_active_regions(int nid, unsigned long start_pfn,
+							unsigned long end_pfn)
+{
+	unsigned long ei_startpfn;
+	unsigned long ei_endpfn;
+	int i;
+
+	for (i = 0; i < e820.nr_map; i++)
+		if (e820_find_active_region(&e820.map[i],
+					    start_pfn, end_pfn,
+					    &ei_startpfn, &ei_endpfn))
+			add_active_range(nid, ei_startpfn, ei_endpfn);
+}
+
+/* 
+ * Add a memory region to the kernel e820 map.
+ */ 
+void __init add_memory_region(unsigned long start, unsigned long size, int type)
+{
+	int x = e820.nr_map;
+
+	if (x == E820MAX) {
+		printk(KERN_ERR "Ooops! Too many entries in the memory map!\n");
+		return;
+	}
+
+	e820.map[x].addr = start;
+	e820.map[x].size = size;
+	e820.map[x].type = type;
+	e820.nr_map++;
+}
+
+/*
+ * Find the hole size (in bytes) in the memory range.
+ * @start: starting address of the memory range to scan
+ * @end: ending address of the memory range to scan
+ */
+unsigned long __init e820_hole_size(unsigned long start, unsigned long end)
+{
+	unsigned long start_pfn = start >> PAGE_SHIFT;
+	unsigned long end_pfn = end >> PAGE_SHIFT;
+	unsigned long ei_startpfn;
+	unsigned long ei_endpfn;
+	unsigned long ram = 0;
+	int i;
+
+	for (i = 0; i < e820.nr_map; i++) {
+		if (e820_find_active_region(&e820.map[i],
+					    start_pfn, end_pfn,
+					    &ei_startpfn, &ei_endpfn))
+			ram += ei_endpfn - ei_startpfn;
+	}
+	return end - start - (ram << PAGE_SHIFT);
+}
+
+void __init e820_print_map(char *who)
+{
+	int i;
+
+	for (i = 0; i < e820.nr_map; i++) {
+		printk(KERN_INFO " %s: %016Lx - %016Lx ", who,
+			(unsigned long long) e820.map[i].addr,
+			(unsigned long long) (e820.map[i].addr + e820.map[i].size));
+		switch (e820.map[i].type) {
+		case E820_RAM:	printk("(usable)\n");
+				break;
+		case E820_RESERVED:
+				printk("(reserved)\n");
+				break;
+		case E820_ACPI:
+				printk("(ACPI data)\n");
+				break;
+		case E820_NVS:
+				printk("(ACPI NVS)\n");
+				break;
+		default:	printk("type %u\n", e820.map[i].type);
+				break;
+		}
+	}
+}
+
+/*
+ * Sanitize the BIOS e820 map.
+ *
+ * Some e820 responses include overlapping entries.  The following 
+ * replaces the original e820 map with a new one, removing overlaps.
+ *
+ */
+static int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map)
+{
+	struct change_member {
+		struct e820entry *pbios; /* pointer to original bios entry */
+		unsigned long long addr; /* address for this change point */
+	};
+	static struct change_member change_point_list[2*E820MAX] __initdata;
+	static struct change_member *change_point[2*E820MAX] __initdata;
+	static struct e820entry *overlap_list[E820MAX] __initdata;
+	static struct e820entry new_bios[E820MAX] __initdata;
+	struct change_member *change_tmp;
+	unsigned long current_type, last_type;
+	unsigned long long last_addr;
+	int chgidx, still_changing;
+	int overlap_entries;
+	int new_bios_entry;
+	int old_nr, new_nr, chg_nr;
+	int i;
+
+	/*
+		Visually we're performing the following (1,2,3,4 = memory types)...
+
+		Sample memory map (w/overlaps):
+		   ____22__________________
+		   ______________________4_
+		   ____1111________________
+		   _44_____________________
+		   11111111________________
+		   ____________________33__
+		   ___________44___________
+		   __________33333_________
+		   ______________22________
+		   ___________________2222_
+		   _________111111111______
+		   _____________________11_
+		   _________________4______
+
+		Sanitized equivalent (no overlap):
+		   1_______________________
+		   _44_____________________
+		   ___1____________________
+		   ____22__________________
+		   ______11________________
+		   _________1______________
+		   __________3_____________
+		   ___________44___________
+		   _____________33_________
+		   _______________2________
+		   ________________1_______
+		   _________________4______
+		   ___________________2____
+		   ____________________33__
+		   ______________________4_
+	*/
+
+	/* if there's only one memory region, don't bother */
+	if (*pnr_map < 2)
+		return -1;
+
+	old_nr = *pnr_map;
+
+	/* bail out if we find any unreasonable addresses in bios map */
+	for (i=0; i<old_nr; i++)
+		if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr)
+			return -1;
+
+	/* create pointers for initial change-point information (for sorting) */
+	for (i=0; i < 2*old_nr; i++)
+		change_point[i] = &change_point_list[i];
+
+	/* record all known change-points (starting and ending addresses),
+	   omitting those that are for empty memory regions */
+	chgidx = 0;
+	for (i=0; i < old_nr; i++)	{
+		if (biosmap[i].size != 0) {
+			change_point[chgidx]->addr = biosmap[i].addr;
+			change_point[chgidx++]->pbios = &biosmap[i];
+			change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size;
+			change_point[chgidx++]->pbios = &biosmap[i];
+		}
+	}
+	chg_nr = chgidx;
+
+	/* sort change-point list by memory addresses (low -> high) */
+	still_changing = 1;
+	while (still_changing)	{
+		still_changing = 0;
+		for (i=1; i < chg_nr; i++)  {
+			/* if <current_addr> > <last_addr>, swap */
+			/* or, if current=<start_addr> & last=<end_addr>, swap */
+			if ((change_point[i]->addr < change_point[i-1]->addr) ||
+				((change_point[i]->addr == change_point[i-1]->addr) &&
+				 (change_point[i]->addr == change_point[i]->pbios->addr) &&
+				 (change_point[i-1]->addr != change_point[i-1]->pbios->addr))
+			   )
+			{
+				change_tmp = change_point[i];
+				change_point[i] = change_point[i-1];
+				change_point[i-1] = change_tmp;
+				still_changing=1;
+			}
+		}
+	}
+
+	/* create a new bios memory map, removing overlaps */
+	overlap_entries=0;	 /* number of entries in the overlap table */
+	new_bios_entry=0;	 /* index for creating new bios map entries */
+	last_type = 0;		 /* start with undefined memory type */
+	last_addr = 0;		 /* start with 0 as last starting address */
+	/* loop through change-points, determining affect on the new bios map */
+	for (chgidx=0; chgidx < chg_nr; chgidx++)
+	{
+		/* keep track of all overlapping bios entries */
+		if (change_point[chgidx]->addr == change_point[chgidx]->pbios->addr)
+		{
+			/* add map entry to overlap list (> 1 entry implies an overlap) */
+			overlap_list[overlap_entries++]=change_point[chgidx]->pbios;
+		}
+		else
+		{
+			/* remove entry from list (order independent, so swap with last) */
+			for (i=0; i<overlap_entries; i++)
+			{
+				if (overlap_list[i] == change_point[chgidx]->pbios)
+					overlap_list[i] = overlap_list[overlap_entries-1];
+			}
+			overlap_entries--;
+		}
+		/* if there are overlapping entries, decide which "type" to use */
+		/* (larger value takes precedence -- 1=usable, 2,3,4,4+=unusable) */
+		current_type = 0;
+		for (i=0; i<overlap_entries; i++)
+			if (overlap_list[i]->type > current_type)
+				current_type = overlap_list[i]->type;
+		/* continue building up new bios map based on this information */
+		if (current_type != last_type)	{
+			if (last_type != 0)	 {
+				new_bios[new_bios_entry].size =
+					change_point[chgidx]->addr - last_addr;
+				/* move forward only if the new size was non-zero */
+				if (new_bios[new_bios_entry].size != 0)
+					if (++new_bios_entry >= E820MAX)
+						break; 	/* no more space left for new bios entries */
+			}
+			if (current_type != 0)	{
+				new_bios[new_bios_entry].addr = change_point[chgidx]->addr;
+				new_bios[new_bios_entry].type = current_type;
+				last_addr=change_point[chgidx]->addr;
+			}
+			last_type = current_type;
+		}
+	}
+	new_nr = new_bios_entry;   /* retain count for new bios entries */
+
+	/* copy new bios mapping into original location */
+	memcpy(biosmap, new_bios, new_nr*sizeof(struct e820entry));
+	*pnr_map = new_nr;
+
+	return 0;
+}
+
+/*
+ * Copy the BIOS e820 map into a safe place.
+ *
+ * Sanity-check it while we're at it..
+ *
+ * If we're lucky and live on a modern system, the setup code
+ * will have given us a memory map that we can use to properly
+ * set up memory.  If we aren't, we'll fake a memory map.
+ */
+static int __init copy_e820_map(struct e820entry * biosmap, int nr_map)
+{
+	/* Only one memory region (or negative)? Ignore it */
+	if (nr_map < 2)
+		return -1;
+
+	do {
+		unsigned long start = biosmap->addr;
+		unsigned long size = biosmap->size;
+		unsigned long end = start + size;
+		unsigned long type = biosmap->type;
+
+		/* Overflow in 64 bits? Ignore the memory map. */
+		if (start > end)
+			return -1;
+
+		add_memory_region(start, size, type);
+	} while (biosmap++,--nr_map);
+	return 0;
+}
+
+void early_panic(char *msg)
+{
+	early_printk(msg);
+	panic(msg);
+}
+
+void __init setup_memory_region(void)
+{
+	/*
+	 * Try to copy the BIOS-supplied E820-map.
+	 *
+	 * Otherwise fake a memory map; one section from 0k->640k,
+	 * the next section from 1mb->appropriate_mem_k
+	 */
+	sanitize_e820_map(E820_MAP, &E820_MAP_NR);
+	if (copy_e820_map(E820_MAP, E820_MAP_NR) < 0)
+		early_panic("Cannot find a valid memory map");
+	printk(KERN_INFO "BIOS-provided physical RAM map:\n");
+	e820_print_map("BIOS-e820");
+}
+
+static int __init parse_memopt(char *p)
+{
+	if (!p)
+		return -EINVAL;
+	end_user_pfn = memparse(p, &p);
+	end_user_pfn >>= PAGE_SHIFT;	
+	return 0;
+} 
+early_param("mem", parse_memopt);
+
+static int userdef __initdata;
+
+static int __init parse_memmap_opt(char *p)
+{
+	char *oldp;
+	unsigned long long start_at, mem_size;
+
+	if (!strcmp(p, "exactmap")) {
+#ifdef CONFIG_CRASH_DUMP
+		/* If we are doing a crash dump, we
+		 * still need to know the real mem
+		 * size before original memory map is
+		 * reset.
+		 */
+		e820_register_active_regions(0, 0, -1UL);
+		saved_max_pfn = e820_end_of_ram();
+		remove_all_active_ranges();
+#endif
+		end_pfn_map = 0;
+		e820.nr_map = 0;
+		userdef = 1;
+		return 0;
+	}
+
+	oldp = p;
+	mem_size = memparse(p, &p);
+	if (p == oldp)
+		return -EINVAL;
+	if (*p == '@') {
+		start_at = memparse(p+1, &p);
+		add_memory_region(start_at, mem_size, E820_RAM);
+	} else if (*p == '#') {
+		start_at = memparse(p+1, &p);
+		add_memory_region(start_at, mem_size, E820_ACPI);
+	} else if (*p == '$') {
+		start_at = memparse(p+1, &p);
+		add_memory_region(start_at, mem_size, E820_RESERVED);
+	} else {
+		end_user_pfn = (mem_size >> PAGE_SHIFT);
+	}
+	return *p == '\0' ? 0 : -EINVAL;
+}
+early_param("memmap", parse_memmap_opt);
+
+void __init finish_e820_parsing(void)
+{
+	if (userdef) {
+		printk(KERN_INFO "user-defined physical RAM map:\n");
+		e820_print_map("user");
+	}
+}
+
+unsigned long pci_mem_start = 0xaeedbabe;
+EXPORT_SYMBOL(pci_mem_start);
+
+/*
+ * Search for the biggest gap in the low 32 bits of the e820
+ * memory space.  We pass this space to PCI to assign MMIO resources
+ * for hotplug or unconfigured devices in.
+ * Hopefully the BIOS let enough space left.
+ */
+__init void e820_setup_gap(void)
+{
+	unsigned long gapstart, gapsize, round;
+	unsigned long last;
+	int i;
+	int found = 0;
+
+	last = 0x100000000ull;
+	gapstart = 0x10000000;
+	gapsize = 0x400000;
+	i = e820.nr_map;
+	while (--i >= 0) {
+		unsigned long long start = e820.map[i].addr;
+		unsigned long long end = start + e820.map[i].size;
+
+		/*
+		 * Since "last" is at most 4GB, we know we'll
+		 * fit in 32 bits if this condition is true
+		 */
+		if (last > end) {
+			unsigned long gap = last - end;
+
+			if (gap > gapsize) {
+				gapsize = gap;
+				gapstart = end;
+				found = 1;
+			}
+		}
+		if (start < last)
+			last = start;
+	}
+
+	if (!found) {
+		gapstart = (end_pfn << PAGE_SHIFT) + 1024*1024;
+		printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit address range\n"
+		       KERN_ERR "PCI: Unassigned devices with 32bit resource registers may break!\n");
+	}
+
+	/*
+	 * See how much we want to round up: start off with
+	 * rounding to the next 1MB area.
+	 */
+	round = 0x100000;
+	while ((gapsize >> 4) > round)
+		round += round;
+	/* Fun with two's complement */
+	pci_mem_start = (gapstart + round) & -round;
+
+	printk(KERN_INFO "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n",
+		pci_mem_start, gapstart, gapsize);
+}

arch/x86/kernel/early-quirks_64.c

@@ -0,0 +1,127 @@
+/* Various workarounds for chipset bugs.
+   This code runs very early and can't use the regular PCI subsystem
+   The entries are keyed to PCI bridges which usually identify chipsets
+   uniquely.
+   This is only for whole classes of chipsets with specific problems which
+   need early invasive action (e.g. before the timers are initialized).
+   Most PCI device specific workarounds can be done later and should be
+   in standard PCI quirks
+   Mainboard specific bugs should be handled by DMI entries.
+   CPU specific bugs in setup.c */
+
+#include <linux/pci.h>
+#include <linux/acpi.h>
+#include <linux/pci_ids.h>
+#include <asm/pci-direct.h>
+#include <asm/proto.h>
+#include <asm/iommu.h>
+#include <asm/dma.h>
+
+static void __init via_bugs(void)
+{
+#ifdef CONFIG_IOMMU
+	if ((end_pfn > MAX_DMA32_PFN ||  force_iommu) &&
+	    !iommu_aperture_allowed) {
+		printk(KERN_INFO
+  "Looks like a VIA chipset. Disabling IOMMU. Override with iommu=allowed\n");
+		iommu_aperture_disabled = 1;
+	}
+#endif
+}
+
+#ifdef CONFIG_ACPI
+
+static int __init nvidia_hpet_check(struct acpi_table_header *header)
+{
+	return 0;
+}
+#endif
+
+static void __init nvidia_bugs(void)
+{
+#ifdef CONFIG_ACPI
+	/*
+	 * All timer overrides on Nvidia are
+	 * wrong unless HPET is enabled.
+	 * Unfortunately that's not true on many Asus boards.
+	 * We don't know yet how to detect this automatically, but
+	 * at least allow a command line override.
+	 */
+	if (acpi_use_timer_override)
+		return;
+
+	if (acpi_table_parse(ACPI_SIG_HPET, nvidia_hpet_check)) {
+		acpi_skip_timer_override = 1;
+		printk(KERN_INFO "Nvidia board "
+		       "detected. Ignoring ACPI "
+		       "timer override.\n");
+		printk(KERN_INFO "If you got timer trouble "
+			"try acpi_use_timer_override\n");
+	}
+#endif
+	/* RED-PEN skip them on mptables too? */
+
+}
+
+static void __init ati_bugs(void)
+{
+	if (timer_over_8254 == 1) {
+		timer_over_8254 = 0;
+		printk(KERN_INFO
+	 	"ATI board detected. Disabling timer routing over 8254.\n");
+	}
+}
+
+struct chipset {
+	u16 vendor;
+	void (*f)(void);
+};
+
+static struct chipset early_qrk[] __initdata = {
+	{ PCI_VENDOR_ID_NVIDIA, nvidia_bugs },
+	{ PCI_VENDOR_ID_VIA, via_bugs },
+	{ PCI_VENDOR_ID_ATI, ati_bugs },
+	{}
+};
+
+void __init early_quirks(void)
+{
+	int num, slot, func;
+
+	if (!early_pci_allowed())
+		return;
+
+	/* Poor man's PCI discovery */
+	for (num = 0; num < 32; num++) {
+		for (slot = 0; slot < 32; slot++) {
+			for (func = 0; func < 8; func++) {
+				u32 class;
+				u32 vendor;
+				u8 type;
+				int i;
+				class = read_pci_config(num,slot,func,
+							PCI_CLASS_REVISION);
+				if (class == 0xffffffff)
+					break;
+
+		       		if ((class >> 16) != PCI_CLASS_BRIDGE_PCI)
+					continue;
+
+				vendor = read_pci_config(num, slot, func,
+							 PCI_VENDOR_ID);
+				vendor &= 0xffff;
+
+				for (i = 0; early_qrk[i].f; i++)
+					if (early_qrk[i].vendor == vendor) {
+						early_qrk[i].f();
+						return;
+					}
+
+				type = read_pci_config_byte(num, slot, func,
+							    PCI_HEADER_TYPE);
+				if (!(type & 0x80))
+					break;
+			}
+		}
+	}
+}

arch/x86/kernel/early_printk.c

@@ -1,2 +1,259 @@
+#include <linux/console.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/string.h>
+#include <linux/screen_info.h>
+#include <asm/io.h>
+#include <asm/processor.h>
+#include <asm/fcntl.h>
+#include <xen/hvc-console.h>
 
-#include "../../x86_64/kernel/early_printk.c"
+/* Simple VGA output */
+
+#ifdef __i386__
+#include <asm/setup.h>
+#else
+#include <asm/bootsetup.h>
+#endif
+#define VGABASE		(__ISA_IO_base + 0xb8000)
+
+static int max_ypos = 25, max_xpos = 80;
+static int current_ypos = 25, current_xpos = 0;
+
+static void early_vga_write(struct console *con, const char *str, unsigned n)
+{
+	char c;
+	int  i, k, j;
+
+	while ((c = *str++) != '\0' && n-- > 0) {
+		if (current_ypos >= max_ypos) {
+			/* scroll 1 line up */
+			for (k = 1, j = 0; k < max_ypos; k++, j++) {
+				for (i = 0; i < max_xpos; i++) {
+					writew(readw(VGABASE+2*(max_xpos*k+i)),
+					       VGABASE + 2*(max_xpos*j + i));
+				}
+			}
+			for (i = 0; i < max_xpos; i++)
+				writew(0x720, VGABASE + 2*(max_xpos*j + i));
+			current_ypos = max_ypos-1;
+		}
+		if (c == '\n') {
+			current_xpos = 0;
+			current_ypos++;
+		} else if (c != '\r')  {
+			writew(((0x7 << 8) | (unsigned short) c),
+			       VGABASE + 2*(max_xpos*current_ypos +
+						current_xpos++));
+			if (current_xpos >= max_xpos) {
+				current_xpos = 0;
+				current_ypos++;
+			}
+		}
+	}
+}
+
+static struct console early_vga_console = {
+	.name =		"earlyvga",
+	.write =	early_vga_write,
+	.flags =	CON_PRINTBUFFER,
+	.index =	-1,
+};
+
+/* Serial functions loosely based on a similar package from Klaus P. Gerlicher */
+
+static int early_serial_base = 0x3f8;  /* ttyS0 */
+
+#define XMTRDY          0x20
+
+#define DLAB		0x80
+
+#define TXR             0       /*  Transmit register (WRITE) */
+#define RXR             0       /*  Receive register  (READ)  */
+#define IER             1       /*  Interrupt Enable          */
+#define IIR             2       /*  Interrupt ID              */
+#define FCR             2       /*  FIFO control              */
+#define LCR             3       /*  Line control              */
+#define MCR             4       /*  Modem control             */
+#define LSR             5       /*  Line Status               */
+#define MSR             6       /*  Modem Status              */
+#define DLL             0       /*  Divisor Latch Low         */
+#define DLH             1       /*  Divisor latch High        */
+
+static int early_serial_putc(unsigned char ch)
+{
+	unsigned timeout = 0xffff;
+	while ((inb(early_serial_base + LSR) & XMTRDY) == 0 && --timeout)
+		cpu_relax();
+	outb(ch, early_serial_base + TXR);
+	return timeout ? 0 : -1;
+}
+
+static void early_serial_write(struct console *con, const char *s, unsigned n)
+{
+	while (*s && n-- > 0) {
+		if (*s == '\n')
+			early_serial_putc('\r');
+		early_serial_putc(*s);
+		s++;
+	}
+}
+
+#define DEFAULT_BAUD 9600
+
+static __init void early_serial_init(char *s)
+{
+	unsigned char c;
+	unsigned divisor;
+	unsigned baud = DEFAULT_BAUD;
+	char *e;
+
+	if (*s == ',')
+		++s;
+
+	if (*s) {
+		unsigned port;
+		if (!strncmp(s,"0x",2)) {
+			early_serial_base = simple_strtoul(s, &e, 16);
+		} else {
+			static int bases[] = { 0x3f8, 0x2f8 };
+
+			if (!strncmp(s,"ttyS",4))
+				s += 4;
+			port = simple_strtoul(s, &e, 10);
+			if (port > 1 || s == e)
+				port = 0;
+			early_serial_base = bases[port];
+		}
+		s += strcspn(s, ",");
+		if (*s == ',')
+			s++;
+	}
+
+	outb(0x3, early_serial_base + LCR);	/* 8n1 */
+	outb(0, early_serial_base + IER);	/* no interrupt */
+	outb(0, early_serial_base + FCR);	/* no fifo */
+	outb(0x3, early_serial_base + MCR);	/* DTR + RTS */
+
+	if (*s) {
+		baud = simple_strtoul(s, &e, 0);
+		if (baud == 0 || s == e)
+			baud = DEFAULT_BAUD;
+	}
+
+	divisor = 115200 / baud;
+	c = inb(early_serial_base + LCR);
+	outb(c | DLAB, early_serial_base + LCR);
+	outb(divisor & 0xff, early_serial_base + DLL);
+	outb((divisor >> 8) & 0xff, early_serial_base + DLH);
+	outb(c & ~DLAB, early_serial_base + LCR);
+}
+
+static struct console early_serial_console = {
+	.name =		"earlyser",
+	.write =	early_serial_write,
+	.flags =	CON_PRINTBUFFER,
+	.index =	-1,
+};
+
+/* Console interface to a host file on AMD's SimNow! */
+
+static int simnow_fd;
+
+enum {
+	MAGIC1 = 0xBACCD00A,
+	MAGIC2 = 0xCA110000,
+	XOPEN = 5,
+	XWRITE = 4,
+};
+
+static noinline long simnow(long cmd, long a, long b, long c)
+{
+	long ret;
+	asm volatile("cpuid" :
+		     "=a" (ret) :
+		     "b" (a), "c" (b), "d" (c), "0" (MAGIC1), "D" (cmd + MAGIC2));
+	return ret;
+}
+
+static void __init simnow_init(char *str)
+{
+	char *fn = "klog";
+	if (*str == '=')
+		fn = ++str;
+	/* error ignored */
+	simnow_fd = simnow(XOPEN, (unsigned long)fn, O_WRONLY|O_APPEND|O_CREAT, 0644);
+}
+
+static void simnow_write(struct console *con, const char *s, unsigned n)
+{
+	simnow(XWRITE, simnow_fd, (unsigned long)s, n);
+}
+
+static struct console simnow_console = {
+	.name =		"simnow",
+	.write =	simnow_write,
+	.flags =	CON_PRINTBUFFER,
+	.index =	-1,
+};
+
+/* Direct interface for emergencies */
+struct console *early_console = &early_vga_console;
+static int early_console_initialized = 0;
+
+void early_printk(const char *fmt, ...)
+{
+	char buf[512];
+	int n;
+	va_list ap;
+
+	va_start(ap,fmt);
+	n = vscnprintf(buf,512,fmt,ap);
+	early_console->write(early_console,buf,n);
+	va_end(ap);
+}
+
+static int __initdata keep_early;
+
+static int __init setup_early_printk(char *buf)
+{
+	if (!buf)
+		return 0;
+
+	if (early_console_initialized)
+		return 0;
+	early_console_initialized = 1;
+
+	if (strstr(buf, "keep"))
+		keep_early = 1;
+
+	if (!strncmp(buf, "serial", 6)) {
+		early_serial_init(buf + 6);
+		early_console = &early_serial_console;
+	} else if (!strncmp(buf, "ttyS", 4)) {
+		early_serial_init(buf);
+		early_console = &early_serial_console;
+	} else if (!strncmp(buf, "vga", 3)
+	           && SCREEN_INFO.orig_video_isVGA == 1) {
+		max_xpos = SCREEN_INFO.orig_video_cols;
+		max_ypos = SCREEN_INFO.orig_video_lines;
+		current_ypos = SCREEN_INFO.orig_y;
+		early_console = &early_vga_console;
+ 	} else if (!strncmp(buf, "simnow", 6)) {
+ 		simnow_init(buf + 6);
+ 		early_console = &simnow_console;
+ 		keep_early = 1;
+#ifdef CONFIG_HVC_XEN
+	} else if (!strncmp(buf, "xen", 3)) {
+		early_console = &xenboot_console;
+#endif
+	}
+
+	if (keep_early)
+		early_console->flags &= ~CON_BOOT;
+	else
+		early_console->flags |= CON_BOOT;
+	register_console(early_console);
+	return 0;
+}
+early_param("earlyprintk", setup_early_printk);

arch/x86/kernel/genapic_64.c

@@ -0,0 +1,66 @@
+/*
+ * Copyright 2004 James Cleverdon, IBM.
+ * Subject to the GNU Public License, v.2
+ *
+ * Generic APIC sub-arch probe layer.
+ *
+ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
+ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
+ * James Cleverdon.
+ */
+#include <linux/threads.h>
+#include <linux/cpumask.h>
+#include <linux/string.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/ctype.h>
+#include <linux/init.h>
+
+#include <asm/smp.h>
+#include <asm/ipi.h>
+#include <asm/genapic.h>
+
+#ifdef CONFIG_ACPI
+#include <acpi/acpi_bus.h>
+#endif
+
+/* which logical CPU number maps to which CPU (physical APIC ID) */
+u8 x86_cpu_to_apicid[NR_CPUS] __read_mostly
+					= { [0 ... NR_CPUS-1] = BAD_APICID };
+EXPORT_SYMBOL(x86_cpu_to_apicid);
+
+u8 x86_cpu_to_log_apicid[NR_CPUS]	= { [0 ... NR_CPUS-1] = BAD_APICID };
+
+struct genapic __read_mostly *genapic = &apic_flat;
+
+/*
+ * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode.
+ */
+void __init setup_apic_routing(void)
+{
+#ifdef CONFIG_ACPI
+	/*
+	 * Quirk: some x86_64 machines can only use physical APIC mode
+	 * regardless of how many processors are present (x86_64 ES7000
+	 * is an example).
+	 */
+	if (acpi_gbl_FADT.header.revision > FADT2_REVISION_ID &&
+			(acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL))
+		genapic = &apic_physflat;
+	else
+#endif
+
+	if (cpus_weight(cpu_possible_map) <= 8)
+		genapic = &apic_flat;
+	else
+		genapic = &apic_physflat;
+
+	printk(KERN_INFO "Setting APIC routing to %s\n", genapic->name);
+}
+
+/* Same for both flat and physical. */
+
+void send_IPI_self(int vector)
+{
+	__send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
+}

arch/x86/kernel/genapic_flat_64.c

@@ -0,0 +1,194 @@
+/*
+ * Copyright 2004 James Cleverdon, IBM.
+ * Subject to the GNU Public License, v.2
+ *
+ * Flat APIC subarch code.
+ *
+ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
+ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
+ * James Cleverdon.
+ */
+#include <linux/errno.h>
+#include <linux/threads.h>
+#include <linux/cpumask.h>
+#include <linux/string.h>
+#include <linux/kernel.h>
+#include <linux/ctype.h>
+#include <linux/init.h>
+#include <asm/smp.h>
+#include <asm/ipi.h>
+#include <asm/genapic.h>
+
+static cpumask_t flat_target_cpus(void)
+{
+	return cpu_online_map;
+}
+
+static cpumask_t flat_vector_allocation_domain(int cpu)
+{
+	/* Careful. Some cpus do not strictly honor the set of cpus
+	 * specified in the interrupt destination when using lowest
+	 * priority interrupt delivery mode.
+	 *
+	 * In particular there was a hyperthreading cpu observed to
+	 * deliver interrupts to the wrong hyperthread when only one
+	 * hyperthread was specified in the interrupt desitination.
+	 */
+	cpumask_t domain = { { [0] = APIC_ALL_CPUS, } };
+	return domain;
+}
+
+/*
+ * Set up the logical destination ID.
+ *
+ * Intel recommends to set DFR, LDR and TPR before enabling
+ * an APIC.  See e.g. "AP-388 82489DX User's Manual" (Intel
+ * document number 292116).  So here it goes...
+ */
+static void flat_init_apic_ldr(void)
+{
+	unsigned long val;
+	unsigned long num, id;
+
+	num = smp_processor_id();
+	id = 1UL << num;
+	x86_cpu_to_log_apicid[num] = id;
+	apic_write(APIC_DFR, APIC_DFR_FLAT);
+	val = apic_read(APIC_LDR) & ~APIC_LDR_MASK;
+	val |= SET_APIC_LOGICAL_ID(id);
+	apic_write(APIC_LDR, val);
+}
+
+static void flat_send_IPI_mask(cpumask_t cpumask, int vector)
+{
+	unsigned long mask = cpus_addr(cpumask)[0];
+	unsigned long flags;
+
+	local_irq_save(flags);
+	__send_IPI_dest_field(mask, vector, APIC_DEST_LOGICAL);
+	local_irq_restore(flags);
+}
+
+static void flat_send_IPI_allbutself(int vector)
+{
+#ifdef	CONFIG_HOTPLUG_CPU
+	int hotplug = 1;
+#else
+	int hotplug = 0;
+#endif
+	if (hotplug || vector == NMI_VECTOR) {
+		cpumask_t allbutme = cpu_online_map;
+
+		cpu_clear(smp_processor_id(), allbutme);
+
+		if (!cpus_empty(allbutme))
+			flat_send_IPI_mask(allbutme, vector);
+	} else if (num_online_cpus() > 1) {
+		__send_IPI_shortcut(APIC_DEST_ALLBUT, vector,APIC_DEST_LOGICAL);
+	}
+}
+
+static void flat_send_IPI_all(int vector)
+{
+	if (vector == NMI_VECTOR)
+		flat_send_IPI_mask(cpu_online_map, vector);
+	else
+		__send_IPI_shortcut(APIC_DEST_ALLINC, vector, APIC_DEST_LOGICAL);
+}
+
+static int flat_apic_id_registered(void)
+{
+	return physid_isset(GET_APIC_ID(apic_read(APIC_ID)), phys_cpu_present_map);
+}
+
+static unsigned int flat_cpu_mask_to_apicid(cpumask_t cpumask)
+{
+	return cpus_addr(cpumask)[0] & APIC_ALL_CPUS;
+}
+
+static unsigned int phys_pkg_id(int index_msb)
+{
+	return hard_smp_processor_id() >> index_msb;
+}
+
+struct genapic apic_flat =  {
+	.name = "flat",
+	.int_delivery_mode = dest_LowestPrio,
+	.int_dest_mode = (APIC_DEST_LOGICAL != 0),
+	.target_cpus = flat_target_cpus,
+	.vector_allocation_domain = flat_vector_allocation_domain,
+	.apic_id_registered = flat_apic_id_registered,
+	.init_apic_ldr = flat_init_apic_ldr,
+	.send_IPI_all = flat_send_IPI_all,
+	.send_IPI_allbutself = flat_send_IPI_allbutself,
+	.send_IPI_mask = flat_send_IPI_mask,
+	.cpu_mask_to_apicid = flat_cpu_mask_to_apicid,
+	.phys_pkg_id = phys_pkg_id,
+};
+
+/*
+ * Physflat mode is used when there are more than 8 CPUs on a AMD system.
+ * We cannot use logical delivery in this case because the mask
+ * overflows, so use physical mode.
+ */
+
+static cpumask_t physflat_target_cpus(void)
+{
+	return cpu_online_map;
+}
+
+static cpumask_t physflat_vector_allocation_domain(int cpu)
+{
+	cpumask_t domain = CPU_MASK_NONE;
+	cpu_set(cpu, domain);
+	return domain;
+}
+
+
+static void physflat_send_IPI_mask(cpumask_t cpumask, int vector)
+{
+	send_IPI_mask_sequence(cpumask, vector);
+}
+
+static void physflat_send_IPI_allbutself(int vector)
+{
+	cpumask_t allbutme = cpu_online_map;
+
+	cpu_clear(smp_processor_id(), allbutme);
+	physflat_send_IPI_mask(allbutme, vector);
+}
+
+static void physflat_send_IPI_all(int vector)
+{
+	physflat_send_IPI_mask(cpu_online_map, vector);
+}
+
+static unsigned int physflat_cpu_mask_to_apicid(cpumask_t cpumask)
+{
+	int cpu;
+
+	/*
+	 * We're using fixed IRQ delivery, can only return one phys APIC ID.
+	 * May as well be the first.
+	 */
+	cpu = first_cpu(cpumask);
+	if ((unsigned)cpu < NR_CPUS)
+		return x86_cpu_to_apicid[cpu];
+	else
+		return BAD_APICID;
+}
+
+struct genapic apic_physflat =  {
+	.name = "physical flat",
+	.int_delivery_mode = dest_Fixed,
+	.int_dest_mode = (APIC_DEST_PHYSICAL != 0),
+	.target_cpus = physflat_target_cpus,
+	.vector_allocation_domain = physflat_vector_allocation_domain,
+	.apic_id_registered = flat_apic_id_registered,
+	.init_apic_ldr = flat_init_apic_ldr,/*not needed, but shouldn't hurt*/
+	.send_IPI_all = physflat_send_IPI_all,
+	.send_IPI_allbutself = physflat_send_IPI_allbutself,
+	.send_IPI_mask = physflat_send_IPI_mask,
+	.cpu_mask_to_apicid = physflat_cpu_mask_to_apicid,
+	.phys_pkg_id = phys_pkg_id,
+};

arch/x86/kernel/head64.c

@@ -0,0 +1,86 @@
+/*
+ *  linux/arch/x86_64/kernel/head64.c -- prepare to run common code
+ *
+ *  Copyright (C) 2000 Andrea Arcangeli <andrea@suse.de> SuSE
+ */
+
+#include <linux/init.h>
+#include <linux/linkage.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/percpu.h>
+
+#include <asm/processor.h>
+#include <asm/proto.h>
+#include <asm/smp.h>
+#include <asm/bootsetup.h>
+#include <asm/setup.h>
+#include <asm/desc.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+#include <asm/sections.h>
+
+static void __init zap_identity_mappings(void)
+{
+	pgd_t *pgd = pgd_offset_k(0UL);
+	pgd_clear(pgd);
+	__flush_tlb();
+}
+
+/* Don't add a printk in there. printk relies on the PDA which is not initialized 
+   yet. */
+static void __init clear_bss(void)
+{
+	memset(__bss_start, 0,
+	       (unsigned long) __bss_stop - (unsigned long) __bss_start);
+}
+
+#define NEW_CL_POINTER		0x228	/* Relative to real mode data */
+#define OLD_CL_MAGIC_ADDR	0x20
+#define OLD_CL_MAGIC            0xA33F
+#define OLD_CL_OFFSET           0x22
+
+static void __init copy_bootdata(char *real_mode_data)
+{
+	unsigned long new_data;
+	char * command_line;
+
+	memcpy(x86_boot_params, real_mode_data, BOOT_PARAM_SIZE);
+	new_data = *(u32 *) (x86_boot_params + NEW_CL_POINTER);
+	if (!new_data) {
+		if (OLD_CL_MAGIC != *(u16 *)(real_mode_data + OLD_CL_MAGIC_ADDR)) {
+			return;
+		}
+		new_data = __pa(real_mode_data) + *(u16 *)(real_mode_data + OLD_CL_OFFSET);
+	}
+	command_line = __va(new_data);
+	memcpy(boot_command_line, command_line, COMMAND_LINE_SIZE);
+}
+
+void __init x86_64_start_kernel(char * real_mode_data)
+{
+	int i;
+
+	/* clear bss before set_intr_gate with early_idt_handler */
+	clear_bss();
+
+	/* Make NULL pointers segfault */
+	zap_identity_mappings();
+
+	for (i = 0; i < IDT_ENTRIES; i++)
+		set_intr_gate(i, early_idt_handler);
+	asm volatile("lidt %0" :: "m" (idt_descr));
+
+	early_printk("Kernel alive\n");
+
+ 	for (i = 0; i < NR_CPUS; i++)
+ 		cpu_pda(i) = &boot_cpu_pda[i];
+
+	pda_init(0);
+	copy_bootdata(__va(real_mode_data));
+#ifdef CONFIG_SMP
+	cpu_set(0, cpu_online_map);
+#endif
+	start_kernel();
+}

arch/x86/kernel/head_64.S

@@ -0,0 +1,416 @@
+/*
+ *  linux/arch/x86_64/kernel/head.S -- start in 32bit and switch to 64bit
+ *
+ *  Copyright (C) 2000 Andrea Arcangeli <andrea@suse.de> SuSE
+ *  Copyright (C) 2000 Pavel Machek <pavel@suse.cz>
+ *  Copyright (C) 2000 Karsten Keil <kkeil@suse.de>
+ *  Copyright (C) 2001,2002 Andi Kleen <ak@suse.de>
+ *  Copyright (C) 2005 Eric Biederman <ebiederm@xmission.com>
+ */
+
+
+#include <linux/linkage.h>
+#include <linux/threads.h>
+#include <linux/init.h>
+#include <asm/desc.h>
+#include <asm/segment.h>
+#include <asm/pgtable.h>
+#include <asm/page.h>
+#include <asm/msr.h>
+#include <asm/cache.h>
+
+/* we are not able to switch in one step to the final KERNEL ADRESS SPACE
+ * because we need identity-mapped pages.
+ *
+ */
+
+	.text
+	.section .text.head
+	.code64
+	.globl startup_64
+startup_64:
+
+	/*
+	 * At this point the CPU runs in 64bit mode CS.L = 1 CS.D = 1,
+	 * and someone has loaded an identity mapped page table
+	 * for us.  These identity mapped page tables map all of the
+	 * kernel pages and possibly all of memory.
+	 *
+	 * %esi holds a physical pointer to real_mode_data.
+	 *
+	 * We come here either directly from a 64bit bootloader, or from
+	 * arch/x86_64/boot/compressed/head.S.
+	 *
+	 * We only come here initially at boot nothing else comes here.
+	 *
+	 * Since we may be loaded at an address different from what we were
+	 * compiled to run at we first fixup the physical addresses in our page
+	 * tables and then reload them.
+	 */
+
+	/* Compute the delta between the address I am compiled to run at and the
+	 * address I am actually running at.
+	 */
+	leaq	_text(%rip), %rbp
+	subq	$_text - __START_KERNEL_map, %rbp
+
+	/* Is the address not 2M aligned? */
+	movq	%rbp, %rax
+	andl	$~LARGE_PAGE_MASK, %eax
+	testl	%eax, %eax
+	jnz	bad_address
+
+	/* Is the address too large? */
+	leaq	_text(%rip), %rdx
+	movq	$PGDIR_SIZE, %rax
+	cmpq	%rax, %rdx
+	jae	bad_address
+
+	/* Fixup the physical addresses in the page table
+	 */
+	addq	%rbp, init_level4_pgt + 0(%rip)
+	addq	%rbp, init_level4_pgt + (258*8)(%rip)
+	addq	%rbp, init_level4_pgt + (511*8)(%rip)
+
+	addq	%rbp, level3_ident_pgt + 0(%rip)
+
+	addq	%rbp, level3_kernel_pgt + (510*8)(%rip)
+	addq	%rbp, level3_kernel_pgt + (511*8)(%rip)
+
+	addq	%rbp, level2_fixmap_pgt + (506*8)(%rip)
+
+	/* Add an Identity mapping if I am above 1G */
+	leaq	_text(%rip), %rdi
+	andq	$LARGE_PAGE_MASK, %rdi
+
+	movq	%rdi, %rax
+	shrq	$PUD_SHIFT, %rax
+	andq	$(PTRS_PER_PUD - 1), %rax
+	jz	ident_complete
+
+	leaq	(level2_spare_pgt - __START_KERNEL_map + _KERNPG_TABLE)(%rbp), %rdx
+	leaq	level3_ident_pgt(%rip), %rbx
+	movq	%rdx, 0(%rbx, %rax, 8)
+
+	movq	%rdi, %rax
+	shrq	$PMD_SHIFT, %rax
+	andq	$(PTRS_PER_PMD - 1), %rax
+	leaq	__PAGE_KERNEL_LARGE_EXEC(%rdi), %rdx
+	leaq	level2_spare_pgt(%rip), %rbx
+	movq	%rdx, 0(%rbx, %rax, 8)
+ident_complete:
+
+	/* Fixup the kernel text+data virtual addresses
+	 */
+	leaq	level2_kernel_pgt(%rip), %rdi
+	leaq	4096(%rdi), %r8
+	/* See if it is a valid page table entry */
+1:	testq	$1, 0(%rdi)
+	jz	2f
+	addq	%rbp, 0(%rdi)
+	/* Go to the next page */
+2:	addq	$8, %rdi
+	cmp	%r8, %rdi
+	jne	1b
+
+	/* Fixup phys_base */
+	addq	%rbp, phys_base(%rip)
+
+#ifdef CONFIG_SMP
+	addq	%rbp, trampoline_level4_pgt + 0(%rip)
+	addq	%rbp, trampoline_level4_pgt + (511*8)(%rip)
+#endif
+#ifdef CONFIG_ACPI_SLEEP
+	addq	%rbp, wakeup_level4_pgt + 0(%rip)
+	addq	%rbp, wakeup_level4_pgt + (511*8)(%rip)
+#endif
+
+	/* Due to ENTRY(), sometimes the empty space gets filled with
+	 * zeros. Better take a jmp than relying on empty space being
+	 * filled with 0x90 (nop)
+	 */
+	jmp secondary_startup_64
+ENTRY(secondary_startup_64)
+	/*
+	 * At this point the CPU runs in 64bit mode CS.L = 1 CS.D = 1,
+	 * and someone has loaded a mapped page table.
+	 *
+	 * %esi holds a physical pointer to real_mode_data.
+	 *
+	 * We come here either from startup_64 (using physical addresses)
+	 * or from trampoline.S (using virtual addresses).
+	 *
+	 * Using virtual addresses from trampoline.S removes the need
+	 * to have any identity mapped pages in the kernel page table
+	 * after the boot processor executes this code.
+	 */
+
+	/* Enable PAE mode and PGE */
+	xorq	%rax, %rax
+	btsq	$5, %rax
+	btsq	$7, %rax
+	movq	%rax, %cr4
+
+	/* Setup early boot stage 4 level pagetables. */
+	movq	$(init_level4_pgt - __START_KERNEL_map), %rax
+	addq	phys_base(%rip), %rax
+	movq	%rax, %cr3
+
+	/* Ensure I am executing from virtual addresses */
+	movq	$1f, %rax
+	jmp	*%rax
+1:
+
+	/* Check if nx is implemented */
+	movl	$0x80000001, %eax
+	cpuid
+	movl	%edx,%edi
+
+	/* Setup EFER (Extended Feature Enable Register) */
+	movl	$MSR_EFER, %ecx
+	rdmsr
+	btsl	$_EFER_SCE, %eax	/* Enable System Call */
+	btl	$20,%edi		/* No Execute supported? */
+	jnc     1f
+	btsl	$_EFER_NX, %eax
+1:	wrmsr				/* Make changes effective */
+
+	/* Setup cr0 */
+#define CR0_PM				1		/* protected mode */
+#define CR0_MP				(1<<1)
+#define CR0_ET				(1<<4)
+#define CR0_NE				(1<<5)
+#define CR0_WP				(1<<16)
+#define CR0_AM				(1<<18)
+#define CR0_PAGING 			(1<<31)
+	movl $CR0_PM|CR0_MP|CR0_ET|CR0_NE|CR0_WP|CR0_AM|CR0_PAGING,%eax
+	/* Make changes effective */
+	movq	%rax, %cr0
+
+	/* Setup a boot time stack */
+	movq init_rsp(%rip),%rsp
+
+	/* zero EFLAGS after setting rsp */
+	pushq $0
+	popfq
+
+	/*
+	 * We must switch to a new descriptor in kernel space for the GDT
+	 * because soon the kernel won't have access anymore to the userspace
+	 * addresses where we're currently running on. We have to do that here
+	 * because in 32bit we couldn't load a 64bit linear address.
+	 */
+	lgdt	cpu_gdt_descr(%rip)
+
+	/* set up data segments. actually 0 would do too */
+	movl $__KERNEL_DS,%eax
+	movl %eax,%ds
+	movl %eax,%ss
+	movl %eax,%es
+
+	/*
+	 * We don't really need to load %fs or %gs, but load them anyway
+	 * to kill any stale realmode selectors.  This allows execution
+	 * under VT hardware.
+	 */
+	movl %eax,%fs
+	movl %eax,%gs
+
+	/* 
+	 * Setup up a dummy PDA. this is just for some early bootup code
+	 * that does in_interrupt() 
+	 */ 
+	movl	$MSR_GS_BASE,%ecx
+	movq	$empty_zero_page,%rax
+	movq    %rax,%rdx
+	shrq	$32,%rdx
+	wrmsr	
+
+	/* esi is pointer to real mode structure with interesting info.
+	   pass it to C */
+	movl	%esi, %edi
+	
+	/* Finally jump to run C code and to be on real kernel address
+	 * Since we are running on identity-mapped space we have to jump
+	 * to the full 64bit address, this is only possible as indirect
+	 * jump.  In addition we need to ensure %cs is set so we make this
+	 * a far return.
+	 */
+	movq	initial_code(%rip),%rax
+	pushq	$0		# fake return address to stop unwinder
+	pushq	$__KERNEL_CS	# set correct cs
+	pushq	%rax		# target address in negative space
+	lretq
+
+	/* SMP bootup changes these two */
+#ifndef CONFIG_HOTPLUG_CPU
+	.pushsection .init.data
+#endif
+	.align	8
+	.globl	initial_code
+initial_code:
+	.quad	x86_64_start_kernel
+#ifndef CONFIG_HOTPLUG_CPU
+	.popsection
+#endif
+	.globl init_rsp
+init_rsp:
+	.quad  init_thread_union+THREAD_SIZE-8
+
+bad_address:
+	jmp bad_address
+
+ENTRY(early_idt_handler)
+	cmpl $2,early_recursion_flag(%rip)
+	jz  1f
+	incl early_recursion_flag(%rip)
+	xorl %eax,%eax
+	movq 8(%rsp),%rsi	# get rip
+	movq (%rsp),%rdx
+	movq %cr2,%rcx
+	leaq early_idt_msg(%rip),%rdi
+	call early_printk
+	cmpl $2,early_recursion_flag(%rip)
+	jz  1f
+	call dump_stack
+#ifdef CONFIG_KALLSYMS	
+	leaq early_idt_ripmsg(%rip),%rdi
+	movq 8(%rsp),%rsi	# get rip again
+	call __print_symbol
+#endif
+1:	hlt
+	jmp 1b
+early_recursion_flag:
+	.long 0
+
+early_idt_msg:
+	.asciz "PANIC: early exception rip %lx error %lx cr2 %lx\n"
+early_idt_ripmsg:
+	.asciz "RIP %s\n"
+
+.balign PAGE_SIZE
+
+#define NEXT_PAGE(name) \
+	.balign	PAGE_SIZE; \
+ENTRY(name)
+
+/* Automate the creation of 1 to 1 mapping pmd entries */
+#define PMDS(START, PERM, COUNT)		\
+	i = 0 ;					\
+	.rept (COUNT) ;				\
+	.quad	(START) + (i << 21) + (PERM) ;	\
+	i = i + 1 ;				\
+	.endr
+
+	/*
+	 * This default setting generates an ident mapping at address 0x100000
+	 * and a mapping for the kernel that precisely maps virtual address
+	 * 0xffffffff80000000 to physical address 0x000000. (always using
+	 * 2Mbyte large pages provided by PAE mode)
+	 */
+NEXT_PAGE(init_level4_pgt)
+	.quad	level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE
+	.fill	257,8,0
+	.quad	level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE
+	.fill	252,8,0
+	/* (2^48-(2*1024*1024*1024))/(2^39) = 511 */
+	.quad	level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE
+
+NEXT_PAGE(level3_ident_pgt)
+	.quad	level2_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE
+	.fill	511,8,0
+
+NEXT_PAGE(level3_kernel_pgt)
+	.fill	510,8,0
+	/* (2^48-(2*1024*1024*1024)-((2^39)*511))/(2^30) = 510 */
+	.quad	level2_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE
+	.quad	level2_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE
+
+NEXT_PAGE(level2_fixmap_pgt)
+	.fill	506,8,0
+	.quad	level1_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE
+	/* 8MB reserved for vsyscalls + a 2MB hole = 4 + 1 entries */
+	.fill	5,8,0
+
+NEXT_PAGE(level1_fixmap_pgt)
+	.fill	512,8,0
+
+NEXT_PAGE(level2_ident_pgt)
+	/* Since I easily can, map the first 1G.
+	 * Don't set NX because code runs from these pages.
+	 */
+	PMDS(0x0000000000000000, __PAGE_KERNEL_LARGE_EXEC, PTRS_PER_PMD)
+
+NEXT_PAGE(level2_kernel_pgt)
+	/* 40MB kernel mapping. The kernel code cannot be bigger than that.
+	   When you change this change KERNEL_TEXT_SIZE in page.h too. */
+	/* (2^48-(2*1024*1024*1024)-((2^39)*511)-((2^30)*510)) = 0 */
+	PMDS(0x0000000000000000, __PAGE_KERNEL_LARGE_EXEC|_PAGE_GLOBAL, KERNEL_TEXT_SIZE/PMD_SIZE)
+	/* Module mapping starts here */
+	.fill	(PTRS_PER_PMD - (KERNEL_TEXT_SIZE/PMD_SIZE)),8,0
+
+NEXT_PAGE(level2_spare_pgt)
+	.fill   512,8,0
+
+#undef PMDS
+#undef NEXT_PAGE
+
+	.data
+	.align 16
+	.globl cpu_gdt_descr
+cpu_gdt_descr:
+	.word	gdt_end-cpu_gdt_table-1
+gdt:
+	.quad	cpu_gdt_table
+#ifdef CONFIG_SMP
+	.rept	NR_CPUS-1
+	.word	0
+	.quad	0
+	.endr
+#endif
+
+ENTRY(phys_base)
+	/* This must match the first entry in level2_kernel_pgt */
+	.quad   0x0000000000000000
+
+/* We need valid kernel segments for data and code in long mode too
+ * IRET will check the segment types  kkeil 2000/10/28
+ * Also sysret mandates a special GDT layout 
+ */
+		 		
+	.section .data.page_aligned, "aw"
+	.align PAGE_SIZE
+
+/* The TLS descriptors are currently at a different place compared to i386.
+   Hopefully nobody expects them at a fixed place (Wine?) */
+	
+ENTRY(cpu_gdt_table)
+	.quad	0x0000000000000000	/* NULL descriptor */
+	.quad	0x00cf9b000000ffff	/* __KERNEL32_CS */
+	.quad	0x00af9b000000ffff	/* __KERNEL_CS */
+	.quad	0x00cf93000000ffff	/* __KERNEL_DS */
+	.quad	0x00cffb000000ffff	/* __USER32_CS */
+	.quad	0x00cff3000000ffff	/* __USER_DS, __USER32_DS  */
+	.quad	0x00affb000000ffff	/* __USER_CS */
+	.quad	0x0			/* unused */
+	.quad	0,0			/* TSS */
+	.quad	0,0			/* LDT */
+	.quad   0,0,0			/* three TLS descriptors */ 
+	.quad	0x0000f40000000000	/* node/CPU stored in limit */
+gdt_end:	
+	/* asm/segment.h:GDT_ENTRIES must match this */	
+	/* This should be a multiple of the cache line size */
+	/* GDTs of other CPUs are now dynamically allocated */
+
+	/* zero the remaining page */
+	.fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0
+
+	.section .bss, "aw", @nobits
+	.align L1_CACHE_BYTES
+ENTRY(idt_table)
+	.skip 256 * 16
+
+	.section .bss.page_aligned, "aw", @nobits
+	.align PAGE_SIZE
+ENTRY(empty_zero_page)
+	.skip PAGE_SIZE

arch/x86/kernel/hpet_64.c

@@ -0,0 +1,493 @@
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/init.h>
+#include <linux/mc146818rtc.h>
+#include <linux/time.h>
+#include <linux/clocksource.h>
+#include <linux/ioport.h>
+#include <linux/acpi.h>
+#include <linux/hpet.h>
+#include <asm/pgtable.h>
+#include <asm/vsyscall.h>
+#include <asm/timex.h>
+#include <asm/hpet.h>
+
+#define HPET_MASK	0xFFFFFFFF
+#define HPET_SHIFT	22
+
+/* FSEC = 10^-15 NSEC = 10^-9 */
+#define FSEC_PER_NSEC	1000000
+
+int nohpet __initdata;
+
+unsigned long hpet_address;
+unsigned long hpet_period;	/* fsecs / HPET clock */
+unsigned long hpet_tick;	/* HPET clocks / interrupt */
+
+int hpet_use_timer;		/* Use counter of hpet for time keeping,
+				 * otherwise PIT
+				 */
+
+#ifdef	CONFIG_HPET
+static __init int late_hpet_init(void)
+{
+	struct hpet_data	hd;
+	unsigned int 		ntimer;
+
+	if (!hpet_address)
+        	return 0;
+
+	memset(&hd, 0, sizeof(hd));
+
+	ntimer = hpet_readl(HPET_ID);
+	ntimer = (ntimer & HPET_ID_NUMBER) >> HPET_ID_NUMBER_SHIFT;
+	ntimer++;
+
+	/*
+	 * Register with driver.
+	 * Timer0 and Timer1 is used by platform.
+	 */
+	hd.hd_phys_address = hpet_address;
+	hd.hd_address = (void __iomem *)fix_to_virt(FIX_HPET_BASE);
+	hd.hd_nirqs = ntimer;
+	hd.hd_flags = HPET_DATA_PLATFORM;
+	hpet_reserve_timer(&hd, 0);
+#ifdef	CONFIG_HPET_EMULATE_RTC
+	hpet_reserve_timer(&hd, 1);
+#endif
+	hd.hd_irq[0] = HPET_LEGACY_8254;
+	hd.hd_irq[1] = HPET_LEGACY_RTC;
+	if (ntimer > 2) {
+		struct hpet		*hpet;
+		struct hpet_timer	*timer;
+		int			i;
+
+		hpet = (struct hpet *) fix_to_virt(FIX_HPET_BASE);
+		timer = &hpet->hpet_timers[2];
+		for (i = 2; i < ntimer; timer++, i++)
+			hd.hd_irq[i] = (timer->hpet_config &
+					Tn_INT_ROUTE_CNF_MASK) >>
+				Tn_INT_ROUTE_CNF_SHIFT;
+
+	}
+
+	hpet_alloc(&hd);
+	return 0;
+}
+fs_initcall(late_hpet_init);
+#endif
+
+int hpet_timer_stop_set_go(unsigned long tick)
+{
+	unsigned int cfg;
+
+/*
+ * Stop the timers and reset the main counter.
+ */
+
+	cfg = hpet_readl(HPET_CFG);
+	cfg &= ~(HPET_CFG_ENABLE | HPET_CFG_LEGACY);
+	hpet_writel(cfg, HPET_CFG);
+	hpet_writel(0, HPET_COUNTER);
+	hpet_writel(0, HPET_COUNTER + 4);
+
+/*
+ * Set up timer 0, as periodic with first interrupt to happen at hpet_tick,
+ * and period also hpet_tick.
+ */
+	if (hpet_use_timer) {
+		hpet_writel(HPET_TN_ENABLE | HPET_TN_PERIODIC | HPET_TN_SETVAL |
+		    HPET_TN_32BIT, HPET_T0_CFG);
+		hpet_writel(hpet_tick, HPET_T0_CMP); /* next interrupt */
+		hpet_writel(hpet_tick, HPET_T0_CMP); /* period */
+		cfg |= HPET_CFG_LEGACY;
+	}
+/*
+ * Go!
+ */
+
+	cfg |= HPET_CFG_ENABLE;
+	hpet_writel(cfg, HPET_CFG);
+
+	return 0;
+}
+
+static cycle_t read_hpet(void)
+{
+	return (cycle_t)hpet_readl(HPET_COUNTER);
+}
+
+static cycle_t __vsyscall_fn vread_hpet(void)
+{
+	return readl((void __iomem *)fix_to_virt(VSYSCALL_HPET) + 0xf0);
+}
+
+struct clocksource clocksource_hpet = {
+	.name		= "hpet",
+	.rating		= 250,
+	.read		= read_hpet,
+	.mask		= (cycle_t)HPET_MASK,
+	.mult		= 0, /* set below */
+	.shift		= HPET_SHIFT,
+	.flags		= CLOCK_SOURCE_IS_CONTINUOUS,
+	.vread		= vread_hpet,
+};
+
+int __init hpet_arch_init(void)
+{
+	unsigned int id;
+	u64 tmp;
+
+	if (!hpet_address)
+		return -1;
+	set_fixmap_nocache(FIX_HPET_BASE, hpet_address);
+	__set_fixmap(VSYSCALL_HPET, hpet_address, PAGE_KERNEL_VSYSCALL_NOCACHE);
+
+/*
+ * Read the period, compute tick and quotient.
+ */
+
+	id = hpet_readl(HPET_ID);
+
+	if (!(id & HPET_ID_VENDOR) || !(id & HPET_ID_NUMBER))
+		return -1;
+
+	hpet_period = hpet_readl(HPET_PERIOD);
+	if (hpet_period < 100000 || hpet_period > 100000000)
+		return -1;
+
+	hpet_tick = (FSEC_PER_TICK + hpet_period / 2) / hpet_period;
+
+	hpet_use_timer = (id & HPET_ID_LEGSUP);
+
+	/*
+	 * hpet period is in femto seconds per cycle
+	 * so we need to convert this to ns/cyc units
+	 * aproximated by mult/2^shift
+	 *
+	 *  fsec/cyc * 1nsec/1000000fsec = nsec/cyc = mult/2^shift
+	 *  fsec/cyc * 1ns/1000000fsec * 2^shift = mult
+	 *  fsec/cyc * 2^shift * 1nsec/1000000fsec = mult
+	 *  (fsec/cyc << shift)/1000000 = mult
+	 *  (hpet_period << shift)/FSEC_PER_NSEC = mult
+	 */
+	tmp = (u64)hpet_period << HPET_SHIFT;
+	do_div(tmp, FSEC_PER_NSEC);
+	clocksource_hpet.mult = (u32)tmp;
+	clocksource_register(&clocksource_hpet);
+
+	return hpet_timer_stop_set_go(hpet_tick);
+}
+
+int hpet_reenable(void)
+{
+	return hpet_timer_stop_set_go(hpet_tick);
+}
+
+/*
+ * calibrate_tsc() calibrates the processor TSC in a very simple way, comparing
+ * it to the HPET timer of known frequency.
+ */
+
+#define TICK_COUNT 100000000
+#define SMI_THRESHOLD 50000
+#define MAX_TRIES  5
+
+/*
+ * Some platforms take periodic SMI interrupts with 5ms duration. Make sure none
+ * occurs between the reads of the hpet & TSC.
+ */
+static void __init read_hpet_tsc(int *hpet, int *tsc)
+{
+	int tsc1, tsc2, hpet1, i;
+
+	for (i = 0; i < MAX_TRIES; i++) {
+		tsc1 = get_cycles_sync();
+		hpet1 = hpet_readl(HPET_COUNTER);
+		tsc2 = get_cycles_sync();
+		if ((tsc2 - tsc1) < SMI_THRESHOLD)
+			break;
+	}
+	*hpet = hpet1;
+	*tsc = tsc2;
+}
+
+unsigned int __init hpet_calibrate_tsc(void)
+{
+	int tsc_start, hpet_start;
+	int tsc_now, hpet_now;
+	unsigned long flags;
+
+	local_irq_save(flags);
+
+	read_hpet_tsc(&hpet_start, &tsc_start);
+
+	do {
+		local_irq_disable();
+		read_hpet_tsc(&hpet_now, &tsc_now);
+		local_irq_restore(flags);
+	} while ((tsc_now - tsc_start) < TICK_COUNT &&
+		(hpet_now - hpet_start) < TICK_COUNT);
+
+	return (tsc_now - tsc_start) * 1000000000L
+		/ ((hpet_now - hpet_start) * hpet_period / 1000);
+}
+
+#ifdef CONFIG_HPET_EMULATE_RTC
+/* HPET in LegacyReplacement Mode eats up RTC interrupt line. When, HPET
+ * is enabled, we support RTC interrupt functionality in software.
+ * RTC has 3 kinds of interrupts:
+ * 1) Update Interrupt - generate an interrupt, every sec, when RTC clock
+ *    is updated
+ * 2) Alarm Interrupt - generate an interrupt at a specific time of day
+ * 3) Periodic Interrupt - generate periodic interrupt, with frequencies
+ *    2Hz-8192Hz (2Hz-64Hz for non-root user) (all freqs in powers of 2)
+ * (1) and (2) above are implemented using polling at a frequency of
+ * 64 Hz. The exact frequency is a tradeoff between accuracy and interrupt
+ * overhead. (DEFAULT_RTC_INT_FREQ)
+ * For (3), we use interrupts at 64Hz or user specified periodic
+ * frequency, whichever is higher.
+ */
+#include <linux/rtc.h>
+
+#define DEFAULT_RTC_INT_FREQ 	64
+#define RTC_NUM_INTS 		1
+
+static unsigned long UIE_on;
+static unsigned long prev_update_sec;
+
+static unsigned long AIE_on;
+static struct rtc_time alarm_time;
+
+static unsigned long PIE_on;
+static unsigned long PIE_freq = DEFAULT_RTC_INT_FREQ;
+static unsigned long PIE_count;
+
+static unsigned long hpet_rtc_int_freq; /* RTC interrupt frequency */
+static unsigned int hpet_t1_cmp; /* cached comparator register */
+
+int is_hpet_enabled(void)
+{
+	return hpet_address != 0;
+}
+
+/*
+ * Timer 1 for RTC, we do not use periodic interrupt feature,
+ * even if HPET supports periodic interrupts on Timer 1.
+ * The reason being, to set up a periodic interrupt in HPET, we need to
+ * stop the main counter. And if we do that everytime someone diables/enables
+ * RTC, we will have adverse effect on main kernel timer running on Timer 0.
+ * So, for the time being, simulate the periodic interrupt in software.
+ *
+ * hpet_rtc_timer_init() is called for the first time and during subsequent
+ * interuppts reinit happens through hpet_rtc_timer_reinit().
+ */
+int hpet_rtc_timer_init(void)
+{
+	unsigned int cfg, cnt;
+	unsigned long flags;
+
+	if (!is_hpet_enabled())
+		return 0;
+	/*
+	 * Set the counter 1 and enable the interrupts.
+	 */
+	if (PIE_on && (PIE_freq > DEFAULT_RTC_INT_FREQ))
+		hpet_rtc_int_freq = PIE_freq;
+	else
+		hpet_rtc_int_freq = DEFAULT_RTC_INT_FREQ;
+
+	local_irq_save(flags);
+
+	cnt = hpet_readl(HPET_COUNTER);
+	cnt += ((hpet_tick*HZ)/hpet_rtc_int_freq);
+	hpet_writel(cnt, HPET_T1_CMP);
+	hpet_t1_cmp = cnt;
+
+	cfg = hpet_readl(HPET_T1_CFG);
+	cfg &= ~HPET_TN_PERIODIC;
+	cfg |= HPET_TN_ENABLE | HPET_TN_32BIT;
+	hpet_writel(cfg, HPET_T1_CFG);
+
+	local_irq_restore(flags);
+
+	return 1;
+}
+
+static void hpet_rtc_timer_reinit(void)
+{
+	unsigned int cfg, cnt, ticks_per_int, lost_ints;
+
+	if (unlikely(!(PIE_on | AIE_on | UIE_on))) {
+		cfg = hpet_readl(HPET_T1_CFG);
+		cfg &= ~HPET_TN_ENABLE;
+		hpet_writel(cfg, HPET_T1_CFG);
+		return;
+	}
+
+	if (PIE_on && (PIE_freq > DEFAULT_RTC_INT_FREQ))
+		hpet_rtc_int_freq = PIE_freq;
+	else
+		hpet_rtc_int_freq = DEFAULT_RTC_INT_FREQ;
+
+	/* It is more accurate to use the comparator value than current count.*/
+	ticks_per_int = hpet_tick * HZ / hpet_rtc_int_freq;
+	hpet_t1_cmp += ticks_per_int;
+	hpet_writel(hpet_t1_cmp, HPET_T1_CMP);
+
+	/*
+	 * If the interrupt handler was delayed too long, the write above tries
+	 * to schedule the next interrupt in the past and the hardware would
+	 * not interrupt until the counter had wrapped around.
+	 * So we have to check that the comparator wasn't set to a past time.
+	 */
+	cnt = hpet_readl(HPET_COUNTER);
+	if (unlikely((int)(cnt - hpet_t1_cmp) > 0)) {
+		lost_ints = (cnt - hpet_t1_cmp) / ticks_per_int + 1;
+		/* Make sure that, even with the time needed to execute
+		 * this code, the next scheduled interrupt has been moved
+		 * back to the future: */
+		lost_ints++;
+
+		hpet_t1_cmp += lost_ints * ticks_per_int;
+		hpet_writel(hpet_t1_cmp, HPET_T1_CMP);
+
+		if (PIE_on)
+			PIE_count += lost_ints;
+
+		if (printk_ratelimit())
+			printk(KERN_WARNING "rtc: lost some interrupts at %ldHz.\n",
+			       hpet_rtc_int_freq);
+	}
+}
+
+/*
+ * The functions below are called from rtc driver.
+ * Return 0 if HPET is not being used.
+ * Otherwise do the necessary changes and return 1.
+ */
+int hpet_mask_rtc_irq_bit(unsigned long bit_mask)
+{
+	if (!is_hpet_enabled())
+		return 0;
+
+	if (bit_mask & RTC_UIE)
+		UIE_on = 0;
+	if (bit_mask & RTC_PIE)
+		PIE_on = 0;
+	if (bit_mask & RTC_AIE)
+		AIE_on = 0;
+
+	return 1;
+}
+
+int hpet_set_rtc_irq_bit(unsigned long bit_mask)
+{
+	int timer_init_reqd = 0;
+
+	if (!is_hpet_enabled())
+		return 0;
+
+	if (!(PIE_on | AIE_on | UIE_on))
+		timer_init_reqd = 1;
+
+	if (bit_mask & RTC_UIE) {
+		UIE_on = 1;
+	}
+	if (bit_mask & RTC_PIE) {
+		PIE_on = 1;
+		PIE_count = 0;
+	}
+	if (bit_mask & RTC_AIE) {
+		AIE_on = 1;
+	}
+
+	if (timer_init_reqd)
+		hpet_rtc_timer_init();
+
+	return 1;
+}
+
+int hpet_set_alarm_time(unsigned char hrs, unsigned char min, unsigned char sec)
+{
+	if (!is_hpet_enabled())
+		return 0;
+
+	alarm_time.tm_hour = hrs;
+	alarm_time.tm_min = min;
+	alarm_time.tm_sec = sec;
+
+	return 1;
+}
+
+int hpet_set_periodic_freq(unsigned long freq)
+{
+	if (!is_hpet_enabled())
+		return 0;
+
+	PIE_freq = freq;
+	PIE_count = 0;
+
+	return 1;
+}
+
+int hpet_rtc_dropped_irq(void)
+{
+	if (!is_hpet_enabled())
+		return 0;
+
+	return 1;
+}
+
+irqreturn_t hpet_rtc_interrupt(int irq, void *dev_id)
+{
+	struct rtc_time curr_time;
+	unsigned long rtc_int_flag = 0;
+	int call_rtc_interrupt = 0;
+
+	hpet_rtc_timer_reinit();
+
+	if (UIE_on | AIE_on) {
+		rtc_get_rtc_time(&curr_time);
+	}
+	if (UIE_on) {
+		if (curr_time.tm_sec != prev_update_sec) {
+			/* Set update int info, call real rtc int routine */
+			call_rtc_interrupt = 1;
+			rtc_int_flag = RTC_UF;
+			prev_update_sec = curr_time.tm_sec;
+		}
+	}
+	if (PIE_on) {
+		PIE_count++;
+		if (PIE_count >= hpet_rtc_int_freq/PIE_freq) {
+			/* Set periodic int info, call real rtc int routine */
+			call_rtc_interrupt = 1;
+			rtc_int_flag |= RTC_PF;
+			PIE_count = 0;
+		}
+	}
+	if (AIE_on) {
+		if ((curr_time.tm_sec == alarm_time.tm_sec) &&
+		    (curr_time.tm_min == alarm_time.tm_min) &&
+		    (curr_time.tm_hour == alarm_time.tm_hour)) {
+			/* Set alarm int info, call real rtc int routine */
+			call_rtc_interrupt = 1;
+			rtc_int_flag |= RTC_AF;
+		}
+	}
+	if (call_rtc_interrupt) {
+		rtc_int_flag |= (RTC_IRQF | (RTC_NUM_INTS << 8));
+		rtc_interrupt(rtc_int_flag, dev_id);
+	}
+	return IRQ_HANDLED;
+}
+#endif
+
+static int __init nohpet_setup(char *s)
+{
+	nohpet = 1;
+	return 1;
+}
+
+__setup("nohpet", nohpet_setup);

arch/x86/kernel/i387_64.c

@@ -0,0 +1,151 @@
+/*
+ *  linux/arch/x86_64/kernel/i387.c
+ *
+ *  Copyright (C) 1994 Linus Torvalds
+ *  Copyright (C) 2002 Andi Kleen, SuSE Labs
+ *
+ *  Pentium III FXSR, SSE support
+ *  General FPU state handling cleanups
+ *	Gareth Hughes <gareth@valinux.com>, May 2000
+ * 
+ *  x86-64 rework 2002 Andi Kleen. 
+ *  Does direct fxsave in and out of user space now for signal handlers.
+ *  All the FSAVE<->FXSAVE conversion code has been moved to the 32bit emulation,
+ *  the 64bit user space sees a FXSAVE frame directly. 
+ */
+
+#include <linux/sched.h>
+#include <linux/init.h>
+#include <asm/processor.h>
+#include <asm/i387.h>
+#include <asm/sigcontext.h>
+#include <asm/user.h>
+#include <asm/ptrace.h>
+#include <asm/uaccess.h>
+
+unsigned int mxcsr_feature_mask __read_mostly = 0xffffffff;
+
+void mxcsr_feature_mask_init(void)
+{
+	unsigned int mask;
+	clts();
+	memset(&current->thread.i387.fxsave, 0, sizeof(struct i387_fxsave_struct));
+	asm volatile("fxsave %0" : : "m" (current->thread.i387.fxsave));
+	mask = current->thread.i387.fxsave.mxcsr_mask;
+	if (mask == 0) mask = 0x0000ffbf;
+	mxcsr_feature_mask &= mask;
+	stts();
+}
+
+/*
+ * Called at bootup to set up the initial FPU state that is later cloned
+ * into all processes.
+ */
+void __cpuinit fpu_init(void)
+{
+	unsigned long oldcr0 = read_cr0();
+	extern void __bad_fxsave_alignment(void);
+		
+	if (offsetof(struct task_struct, thread.i387.fxsave) & 15)
+		__bad_fxsave_alignment();
+	set_in_cr4(X86_CR4_OSFXSR);
+	set_in_cr4(X86_CR4_OSXMMEXCPT);
+
+	write_cr0(oldcr0 & ~((1UL<<3)|(1UL<<2))); /* clear TS and EM */
+
+	mxcsr_feature_mask_init();
+	/* clean state in init */
+	current_thread_info()->status = 0;
+	clear_used_math();
+}
+
+void init_fpu(struct task_struct *child)
+{
+	if (tsk_used_math(child)) {
+		if (child == current)
+			unlazy_fpu(child);
+		return;
+	}	
+	memset(&child->thread.i387.fxsave, 0, sizeof(struct i387_fxsave_struct));
+	child->thread.i387.fxsave.cwd = 0x37f;
+	child->thread.i387.fxsave.mxcsr = 0x1f80;
+	/* only the device not available exception or ptrace can call init_fpu */
+	set_stopped_child_used_math(child);
+}
+
+/*
+ * Signal frame handlers.
+ */
+
+int save_i387(struct _fpstate __user *buf)
+{
+	struct task_struct *tsk = current;
+	int err = 0;
+
+	BUILD_BUG_ON(sizeof(struct user_i387_struct) !=
+			sizeof(tsk->thread.i387.fxsave));
+
+	if ((unsigned long)buf % 16) 
+		printk("save_i387: bad fpstate %p\n",buf); 
+
+	if (!used_math())
+		return 0;
+	clear_used_math(); /* trigger finit */
+	if (task_thread_info(tsk)->status & TS_USEDFPU) {
+		err = save_i387_checking((struct i387_fxsave_struct __user *)buf);
+		if (err) return err;
+		stts();
+		} else {
+		if (__copy_to_user(buf, &tsk->thread.i387.fxsave, 
+				   sizeof(struct i387_fxsave_struct)))
+			return -1;
+	} 
+		return 1;
+}
+
+/*
+ * ptrace request handlers.
+ */
+
+int get_fpregs(struct user_i387_struct __user *buf, struct task_struct *tsk)
+{
+	init_fpu(tsk);
+	return __copy_to_user(buf, &tsk->thread.i387.fxsave,
+			       sizeof(struct user_i387_struct)) ? -EFAULT : 0;
+}
+
+int set_fpregs(struct task_struct *tsk, struct user_i387_struct __user *buf)
+{
+	if (__copy_from_user(&tsk->thread.i387.fxsave, buf, 
+			     sizeof(struct user_i387_struct)))
+		return -EFAULT;
+		return 0;
+}
+
+/*
+ * FPU state for core dumps.
+ */
+
+int dump_fpu( struct pt_regs *regs, struct user_i387_struct *fpu )
+{
+	struct task_struct *tsk = current;
+
+	if (!used_math())
+		return 0;
+
+	unlazy_fpu(tsk);
+	memcpy(fpu, &tsk->thread.i387.fxsave, sizeof(struct user_i387_struct)); 
+	return 1; 
+}
+
+int dump_task_fpu(struct task_struct *tsk, struct user_i387_struct *fpu)
+{
+	int fpvalid = !!tsk_used_math(tsk);
+
+	if (fpvalid) {
+		if (tsk == current)
+			unlazy_fpu(tsk);
+		memcpy(fpu, &tsk->thread.i387.fxsave, sizeof(struct user_i387_struct)); 	
+}
+	return fpvalid;
+}

arch/x86/kernel/i8259_64.c

@@ -0,0 +1,544 @@
+#include <linux/linkage.h>
+#include <linux/errno.h>
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/ioport.h>
+#include <linux/interrupt.h>
+#include <linux/timex.h>
+#include <linux/slab.h>
+#include <linux/random.h>
+#include <linux/init.h>
+#include <linux/kernel_stat.h>
+#include <linux/sysdev.h>
+#include <linux/bitops.h>
+
+#include <asm/acpi.h>
+#include <asm/atomic.h>
+#include <asm/system.h>
+#include <asm/io.h>
+#include <asm/hw_irq.h>
+#include <asm/pgtable.h>
+#include <asm/delay.h>
+#include <asm/desc.h>
+#include <asm/apic.h>
+
+/*
+ * Common place to define all x86 IRQ vectors
+ *
+ * This builds up the IRQ handler stubs using some ugly macros in irq.h
+ *
+ * These macros create the low-level assembly IRQ routines that save
+ * register context and call do_IRQ(). do_IRQ() then does all the
+ * operations that are needed to keep the AT (or SMP IOAPIC)
+ * interrupt-controller happy.
+ */
+
+#define BI(x,y) \
+	BUILD_IRQ(x##y)
+
+#define BUILD_16_IRQS(x) \
+	BI(x,0) BI(x,1) BI(x,2) BI(x,3) \
+	BI(x,4) BI(x,5) BI(x,6) BI(x,7) \
+	BI(x,8) BI(x,9) BI(x,a) BI(x,b) \
+	BI(x,c) BI(x,d) BI(x,e) BI(x,f)
+
+/*
+ * ISA PIC or low IO-APIC triggered (INTA-cycle or APIC) interrupts:
+ * (these are usually mapped to vectors 0x30-0x3f)
+ */
+
+/*
+ * The IO-APIC gives us many more interrupt sources. Most of these 
+ * are unused but an SMP system is supposed to have enough memory ...
+ * sometimes (mostly wrt. hw bugs) we get corrupted vectors all
+ * across the spectrum, so we really want to be prepared to get all
+ * of these. Plus, more powerful systems might have more than 64
+ * IO-APIC registers.
+ *
+ * (these are usually mapped into the 0x30-0xff vector range)
+ */
+				      BUILD_16_IRQS(0x2) BUILD_16_IRQS(0x3)
+BUILD_16_IRQS(0x4) BUILD_16_IRQS(0x5) BUILD_16_IRQS(0x6) BUILD_16_IRQS(0x7)
+BUILD_16_IRQS(0x8) BUILD_16_IRQS(0x9) BUILD_16_IRQS(0xa) BUILD_16_IRQS(0xb)
+BUILD_16_IRQS(0xc) BUILD_16_IRQS(0xd) BUILD_16_IRQS(0xe) BUILD_16_IRQS(0xf)
+
+#undef BUILD_16_IRQS
+#undef BI
+
+
+#define IRQ(x,y) \
+	IRQ##x##y##_interrupt
+
+#define IRQLIST_16(x) \
+	IRQ(x,0), IRQ(x,1), IRQ(x,2), IRQ(x,3), \
+	IRQ(x,4), IRQ(x,5), IRQ(x,6), IRQ(x,7), \
+	IRQ(x,8), IRQ(x,9), IRQ(x,a), IRQ(x,b), \
+	IRQ(x,c), IRQ(x,d), IRQ(x,e), IRQ(x,f)
+
+/* for the irq vectors */
+static void (*interrupt[NR_VECTORS - FIRST_EXTERNAL_VECTOR])(void) = {
+					  IRQLIST_16(0x2), IRQLIST_16(0x3),
+	IRQLIST_16(0x4), IRQLIST_16(0x5), IRQLIST_16(0x6), IRQLIST_16(0x7),
+	IRQLIST_16(0x8), IRQLIST_16(0x9), IRQLIST_16(0xa), IRQLIST_16(0xb),
+	IRQLIST_16(0xc), IRQLIST_16(0xd), IRQLIST_16(0xe), IRQLIST_16(0xf)
+};
+
+#undef IRQ
+#undef IRQLIST_16
+
+/*
+ * This is the 'legacy' 8259A Programmable Interrupt Controller,
+ * present in the majority of PC/AT boxes.
+ * plus some generic x86 specific things if generic specifics makes
+ * any sense at all.
+ * this file should become arch/i386/kernel/irq.c when the old irq.c
+ * moves to arch independent land
+ */
+
+static int i8259A_auto_eoi;
+DEFINE_SPINLOCK(i8259A_lock);
+static void mask_and_ack_8259A(unsigned int);
+
+static struct irq_chip i8259A_chip = {
+	.name		= "XT-PIC",
+	.mask		= disable_8259A_irq,
+	.disable	= disable_8259A_irq,
+	.unmask		= enable_8259A_irq,
+	.mask_ack	= mask_and_ack_8259A,
+};
+
+/*
+ * 8259A PIC functions to handle ISA devices:
+ */
+
+/*
+ * This contains the irq mask for both 8259A irq controllers,
+ */
+static unsigned int cached_irq_mask = 0xffff;
+
+#define __byte(x,y) 	(((unsigned char *)&(y))[x])
+#define cached_21	(__byte(0,cached_irq_mask))
+#define cached_A1	(__byte(1,cached_irq_mask))
+
+/*
+ * Not all IRQs can be routed through the IO-APIC, eg. on certain (older)
+ * boards the timer interrupt is not really connected to any IO-APIC pin,
+ * it's fed to the master 8259A's IR0 line only.
+ *
+ * Any '1' bit in this mask means the IRQ is routed through the IO-APIC.
+ * this 'mixed mode' IRQ handling costs nothing because it's only used
+ * at IRQ setup time.
+ */
+unsigned long io_apic_irqs;
+
+void disable_8259A_irq(unsigned int irq)
+{
+	unsigned int mask = 1 << irq;
+	unsigned long flags;
+
+	spin_lock_irqsave(&i8259A_lock, flags);
+	cached_irq_mask |= mask;
+	if (irq & 8)
+		outb(cached_A1,0xA1);
+	else
+		outb(cached_21,0x21);
+	spin_unlock_irqrestore(&i8259A_lock, flags);
+}
+
+void enable_8259A_irq(unsigned int irq)
+{
+	unsigned int mask = ~(1 << irq);
+	unsigned long flags;
+
+	spin_lock_irqsave(&i8259A_lock, flags);
+	cached_irq_mask &= mask;
+	if (irq & 8)
+		outb(cached_A1,0xA1);
+	else
+		outb(cached_21,0x21);
+	spin_unlock_irqrestore(&i8259A_lock, flags);
+}
+
+int i8259A_irq_pending(unsigned int irq)
+{
+	unsigned int mask = 1<<irq;
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&i8259A_lock, flags);
+	if (irq < 8)
+		ret = inb(0x20) & mask;
+	else
+		ret = inb(0xA0) & (mask >> 8);
+	spin_unlock_irqrestore(&i8259A_lock, flags);
+
+	return ret;
+}
+
+void make_8259A_irq(unsigned int irq)
+{
+	disable_irq_nosync(irq);
+	io_apic_irqs &= ~(1<<irq);
+	set_irq_chip_and_handler_name(irq, &i8259A_chip, handle_level_irq,
+				      "XT");
+	enable_irq(irq);
+}
+
+/*
+ * This function assumes to be called rarely. Switching between
+ * 8259A registers is slow.
+ * This has to be protected by the irq controller spinlock
+ * before being called.
+ */
+static inline int i8259A_irq_real(unsigned int irq)
+{
+	int value;
+	int irqmask = 1<<irq;
+
+	if (irq < 8) {
+		outb(0x0B,0x20);		/* ISR register */
+		value = inb(0x20) & irqmask;
+		outb(0x0A,0x20);		/* back to the IRR register */
+		return value;
+	}
+	outb(0x0B,0xA0);		/* ISR register */
+	value = inb(0xA0) & (irqmask >> 8);
+	outb(0x0A,0xA0);		/* back to the IRR register */
+	return value;
+}
+
+/*
+ * Careful! The 8259A is a fragile beast, it pretty
+ * much _has_ to be done exactly like this (mask it
+ * first, _then_ send the EOI, and the order of EOI
+ * to the two 8259s is important!
+ */
+static void mask_and_ack_8259A(unsigned int irq)
+{
+	unsigned int irqmask = 1 << irq;
+	unsigned long flags;
+
+	spin_lock_irqsave(&i8259A_lock, flags);
+	/*
+	 * Lightweight spurious IRQ detection. We do not want
+	 * to overdo spurious IRQ handling - it's usually a sign
+	 * of hardware problems, so we only do the checks we can
+	 * do without slowing down good hardware unnecessarily.
+	 *
+	 * Note that IRQ7 and IRQ15 (the two spurious IRQs
+	 * usually resulting from the 8259A-1|2 PICs) occur
+	 * even if the IRQ is masked in the 8259A. Thus we
+	 * can check spurious 8259A IRQs without doing the
+	 * quite slow i8259A_irq_real() call for every IRQ.
+	 * This does not cover 100% of spurious interrupts,
+	 * but should be enough to warn the user that there
+	 * is something bad going on ...
+	 */
+	if (cached_irq_mask & irqmask)
+		goto spurious_8259A_irq;
+	cached_irq_mask |= irqmask;
+
+handle_real_irq:
+	if (irq & 8) {
+		inb(0xA1);		/* DUMMY - (do we need this?) */
+		outb(cached_A1,0xA1);
+		outb(0x60+(irq&7),0xA0);/* 'Specific EOI' to slave */
+		outb(0x62,0x20);	/* 'Specific EOI' to master-IRQ2 */
+	} else {
+		inb(0x21);		/* DUMMY - (do we need this?) */
+		outb(cached_21,0x21);
+		outb(0x60+irq,0x20);	/* 'Specific EOI' to master */
+	}
+	spin_unlock_irqrestore(&i8259A_lock, flags);
+	return;
+
+spurious_8259A_irq:
+	/*
+	 * this is the slow path - should happen rarely.
+	 */
+	if (i8259A_irq_real(irq))
+		/*
+		 * oops, the IRQ _is_ in service according to the
+		 * 8259A - not spurious, go handle it.
+		 */
+		goto handle_real_irq;
+
+	{
+		static int spurious_irq_mask;
+		/*
+		 * At this point we can be sure the IRQ is spurious,
+		 * lets ACK and report it. [once per IRQ]
+		 */
+		if (!(spurious_irq_mask & irqmask)) {
+			printk(KERN_DEBUG "spurious 8259A interrupt: IRQ%d.\n", irq);
+			spurious_irq_mask |= irqmask;
+		}
+		atomic_inc(&irq_err_count);
+		/*
+		 * Theoretically we do not have to handle this IRQ,
+		 * but in Linux this does not cause problems and is
+		 * simpler for us.
+		 */
+		goto handle_real_irq;
+	}
+}
+
+void init_8259A(int auto_eoi)
+{
+	unsigned long flags;
+
+	i8259A_auto_eoi = auto_eoi;
+
+	spin_lock_irqsave(&i8259A_lock, flags);
+
+	outb(0xff, 0x21);	/* mask all of 8259A-1 */
+	outb(0xff, 0xA1);	/* mask all of 8259A-2 */
+
+	/*
+	 * outb_p - this has to work on a wide range of PC hardware.
+	 */
+	outb_p(0x11, 0x20);	/* ICW1: select 8259A-1 init */
+	outb_p(IRQ0_VECTOR, 0x21);	/* ICW2: 8259A-1 IR0-7 mapped to 0x30-0x37 */
+	outb_p(0x04, 0x21);	/* 8259A-1 (the master) has a slave on IR2 */
+	if (auto_eoi)
+		outb_p(0x03, 0x21);	/* master does Auto EOI */
+	else
+		outb_p(0x01, 0x21);	/* master expects normal EOI */
+
+	outb_p(0x11, 0xA0);	/* ICW1: select 8259A-2 init */
+	outb_p(IRQ8_VECTOR, 0xA1);	/* ICW2: 8259A-2 IR0-7 mapped to 0x38-0x3f */
+	outb_p(0x02, 0xA1);	/* 8259A-2 is a slave on master's IR2 */
+	outb_p(0x01, 0xA1);	/* (slave's support for AEOI in flat mode
+				    is to be investigated) */
+
+	if (auto_eoi)
+		/*
+		 * in AEOI mode we just have to mask the interrupt
+		 * when acking.
+		 */
+		i8259A_chip.mask_ack = disable_8259A_irq;
+	else
+		i8259A_chip.mask_ack = mask_and_ack_8259A;
+
+	udelay(100);		/* wait for 8259A to initialize */
+
+	outb(cached_21, 0x21);	/* restore master IRQ mask */
+	outb(cached_A1, 0xA1);	/* restore slave IRQ mask */
+
+	spin_unlock_irqrestore(&i8259A_lock, flags);
+}
+
+static char irq_trigger[2];
+/**
+ * ELCR registers (0x4d0, 0x4d1) control edge/level of IRQ
+ */
+static void restore_ELCR(char *trigger)
+{
+	outb(trigger[0], 0x4d0);
+	outb(trigger[1], 0x4d1);
+}
+
+static void save_ELCR(char *trigger)
+{
+	/* IRQ 0,1,2,8,13 are marked as reserved */
+	trigger[0] = inb(0x4d0) & 0xF8;
+	trigger[1] = inb(0x4d1) & 0xDE;
+}
+
+static int i8259A_resume(struct sys_device *dev)
+{
+	init_8259A(i8259A_auto_eoi);
+	restore_ELCR(irq_trigger);
+	return 0;
+}
+
+static int i8259A_suspend(struct sys_device *dev, pm_message_t state)
+{
+	save_ELCR(irq_trigger);
+	return 0;
+}
+
+static int i8259A_shutdown(struct sys_device *dev)
+{
+	/* Put the i8259A into a quiescent state that
+	 * the kernel initialization code can get it
+	 * out of.
+	 */
+	outb(0xff, 0x21);	/* mask all of 8259A-1 */
+	outb(0xff, 0xA1);	/* mask all of 8259A-1 */
+	return 0;
+}
+
+static struct sysdev_class i8259_sysdev_class = {
+	set_kset_name("i8259"),
+	.suspend = i8259A_suspend,
+	.resume = i8259A_resume,
+	.shutdown = i8259A_shutdown,
+};
+
+static struct sys_device device_i8259A = {
+	.id	= 0,
+	.cls	= &i8259_sysdev_class,
+};
+
+static int __init i8259A_init_sysfs(void)
+{
+	int error = sysdev_class_register(&i8259_sysdev_class);
+	if (!error)
+		error = sysdev_register(&device_i8259A);
+	return error;
+}
+
+device_initcall(i8259A_init_sysfs);
+
+/*
+ * IRQ2 is cascade interrupt to second interrupt controller
+ */
+
+static struct irqaction irq2 = { no_action, 0, CPU_MASK_NONE, "cascade", NULL, NULL};
+DEFINE_PER_CPU(vector_irq_t, vector_irq) = {
+	[0 ... IRQ0_VECTOR - 1] = -1,
+	[IRQ0_VECTOR] = 0,
+	[IRQ1_VECTOR] = 1,
+	[IRQ2_VECTOR] = 2,
+	[IRQ3_VECTOR] = 3,
+	[IRQ4_VECTOR] = 4,
+	[IRQ5_VECTOR] = 5,
+	[IRQ6_VECTOR] = 6,
+	[IRQ7_VECTOR] = 7,
+	[IRQ8_VECTOR] = 8,
+	[IRQ9_VECTOR] = 9,
+	[IRQ10_VECTOR] = 10,
+	[IRQ11_VECTOR] = 11,
+	[IRQ12_VECTOR] = 12,
+	[IRQ13_VECTOR] = 13,
+	[IRQ14_VECTOR] = 14,
+	[IRQ15_VECTOR] = 15,
+	[IRQ15_VECTOR + 1 ... NR_VECTORS - 1] = -1
+};
+
+void __init init_ISA_irqs (void)
+{
+	int i;
+
+	init_bsp_APIC();
+	init_8259A(0);
+
+	for (i = 0; i < NR_IRQS; i++) {
+		irq_desc[i].status = IRQ_DISABLED;
+		irq_desc[i].action = NULL;
+		irq_desc[i].depth = 1;
+
+		if (i < 16) {
+			/*
+			 * 16 old-style INTA-cycle interrupts:
+			 */
+			set_irq_chip_and_handler_name(i, &i8259A_chip,
+						      handle_level_irq, "XT");
+		} else {
+			/*
+			 * 'high' PCI IRQs filled in on demand
+			 */
+			irq_desc[i].chip = &no_irq_chip;
+		}
+	}
+}
+
+static void setup_timer_hardware(void)
+{
+	outb_p(0x34,0x43);		/* binary, mode 2, LSB/MSB, ch 0 */
+	udelay(10);
+	outb_p(LATCH & 0xff , 0x40);	/* LSB */
+	udelay(10);
+	outb(LATCH >> 8 , 0x40);	/* MSB */
+}
+
+static int timer_resume(struct sys_device *dev)
+{
+	setup_timer_hardware();
+	return 0;
+}
+
+void i8254_timer_resume(void)
+{
+	setup_timer_hardware();
+}
+
+static struct sysdev_class timer_sysclass = {
+	set_kset_name("timer_pit"),
+	.resume		= timer_resume,
+};
+
+static struct sys_device device_timer = {
+	.id		= 0,
+	.cls		= &timer_sysclass,
+};
+
+static int __init init_timer_sysfs(void)
+{
+	int error = sysdev_class_register(&timer_sysclass);
+	if (!error)
+		error = sysdev_register(&device_timer);
+	return error;
+}
+
+device_initcall(init_timer_sysfs);
+
+void __init init_IRQ(void)
+{
+	int i;
+
+	init_ISA_irqs();
+	/*
+	 * Cover the whole vector space, no vector can escape
+	 * us. (some of these will be overridden and become
+	 * 'special' SMP interrupts)
+	 */
+	for (i = 0; i < (NR_VECTORS - FIRST_EXTERNAL_VECTOR); i++) {
+		int vector = FIRST_EXTERNAL_VECTOR + i;
+		if (vector != IA32_SYSCALL_VECTOR)
+			set_intr_gate(vector, interrupt[i]);
+	}
+
+#ifdef CONFIG_SMP
+	/*
+	 * The reschedule interrupt is a CPU-to-CPU reschedule-helper
+	 * IPI, driven by wakeup.
+	 */
+	set_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);
+
+	/* IPI for generic function call */
+	set_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);
+
+	/* self generated IPI for local APIC timer */
+	set_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);
+
+	/*
+	 * Set the clock to HZ Hz, we already have a valid
+	 * vector now:
+	 */
+	setup_timer_hardware();
+
+	if (!acpi_ioapic)
+		setup_irq(2, &irq2);
+}

arch/x86/kernel/init_task_64.c

@@ -0,0 +1,54 @@
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/init.h>
+#include <linux/init_task.h>
+#include <linux/fs.h>
+#include <linux/mqueue.h>
+
+#include <asm/uaccess.h>
+#include <asm/pgtable.h>
+#include <asm/desc.h>
+
+static struct fs_struct init_fs = INIT_FS;
+static struct files_struct init_files = INIT_FILES;
+static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
+static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
+struct mm_struct init_mm = INIT_MM(init_mm);
+
+EXPORT_SYMBOL(init_mm);
+
+/*
+ * Initial task structure.
+ *
+ * We need to make sure that this is 8192-byte aligned due to the
+ * way process stacks are handled. This is done by having a special
+ * "init_task" linker map entry..
+ */
+union thread_union init_thread_union 
+	__attribute__((__section__(".data.init_task"))) =
+		{ INIT_THREAD_INFO(init_task) };
+
+/*
+ * Initial task structure.
+ *
+ * All other task structs will be allocated on slabs in fork.c
+ */
+struct task_struct init_task = INIT_TASK(init_task);
+
+EXPORT_SYMBOL(init_task);
+/*
+ * per-CPU TSS segments. Threads are completely 'soft' on Linux,
+ * no more per-task TSS's. The TSS size is kept cacheline-aligned
+ * so they are allowed to end up in the .data.cacheline_aligned
+ * section. Since TSS's are completely CPU-local, we want them
+ * on exact cacheline boundaries, to eliminate cacheline ping-pong.
+ */ 
+DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS;
+
+/* Copies of the original ist values from the tss are only accessed during
+ * debugging, no special alignment required.
+ */
+DEFINE_PER_CPU(struct orig_ist, orig_ist);
+
+#define ALIGN_TO_4K __attribute__((section(".data.init_task")))

arch/x86/kernel/ioport_64.c

@@ -0,0 +1,119 @@
+/*
+ *	linux/arch/x86_64/kernel/ioport.c
+ *
+ * This contains the io-permission bitmap code - written by obz, with changes
+ * by Linus.
+ */
+
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/capability.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/ioport.h>
+#include <linux/smp.h>
+#include <linux/stddef.h>
+#include <linux/slab.h>
+#include <linux/thread_info.h>
+#include <linux/syscalls.h>
+
+/* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
+static void set_bitmap(unsigned long *bitmap, unsigned int base, unsigned int extent, int new_value)
+{
+	int i;
+		if (new_value)
+		for (i = base; i < base + extent; i++) 
+			__set_bit(i, bitmap); 
+		else
+		for (i = base; i < base + extent; i++) 
+			clear_bit(i, bitmap); 
+}
+
+/*
+ * this changes the io permissions bitmap in the current task.
+ */
+asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on)
+{
+	unsigned int i, max_long, bytes, bytes_updated;
+	struct thread_struct * t = &current->thread;
+	struct tss_struct * tss;
+	unsigned long *bitmap;
+
+	if ((from + num <= from) || (from + num > IO_BITMAP_BITS))
+		return -EINVAL;
+	if (turn_on && !capable(CAP_SYS_RAWIO))
+		return -EPERM;
+
+	/*
+	 * If it's the first ioperm() call in this thread's lifetime, set the
+	 * IO bitmap up. ioperm() is much less timing critical than clone(),
+	 * this is why we delay this operation until now:
+	 */
+	if (!t->io_bitmap_ptr) {
+		bitmap = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
+		if (!bitmap)
+			return -ENOMEM;
+
+		memset(bitmap, 0xff, IO_BITMAP_BYTES);
+		t->io_bitmap_ptr = bitmap;
+		set_thread_flag(TIF_IO_BITMAP);
+	}
+
+	/*
+	 * do it in the per-thread copy and in the TSS ...
+	 *
+	 * Disable preemption via get_cpu() - we must not switch away
+	 * because the ->io_bitmap_max value must match the bitmap
+	 * contents:
+	 */
+	tss = &per_cpu(init_tss, get_cpu());
+
+	set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
+
+	/*
+	 * Search for a (possibly new) maximum. This is simple and stupid,
+	 * to keep it obviously correct:
+	 */
+	max_long = 0;
+	for (i = 0; i < IO_BITMAP_LONGS; i++)
+		if (t->io_bitmap_ptr[i] != ~0UL)
+			max_long = i;
+
+	bytes = (max_long + 1) * sizeof(long);
+	bytes_updated = max(bytes, t->io_bitmap_max);
+
+	t->io_bitmap_max = bytes;
+
+	/* Update the TSS: */
+	memcpy(tss->io_bitmap, t->io_bitmap_ptr, bytes_updated);
+
+	put_cpu();
+
+	return 0;
+}
+
+/*
+ * sys_iopl has to be used when you want to access the IO ports
+ * beyond the 0x3ff range: to get the full 65536 ports bitmapped
+ * you'd need 8kB of bitmaps/process, which is a bit excessive.
+ *
+ * Here we just change the eflags value on the stack: we allow
+ * only the super-user to do it. This depends on the stack-layout
+ * on system-call entry - see also fork() and the signal handling
+ * code.
+ */
+
+asmlinkage long sys_iopl(unsigned int level, struct pt_regs *regs)
+{
+	unsigned int old = (regs->eflags >> 12) & 3;
+
+	if (level > 3)
+		return -EINVAL;
+	/* Trying to gain more privileges? */
+	if (level > old) {
+		if (!capable(CAP_SYS_RAWIO))
+			return -EPERM;
+	}
+	regs->eflags = (regs->eflags &~ X86_EFLAGS_IOPL) | (level << 12);
+	return 0;
+}

arch/x86/kernel/irq_64.c

@@ -0,0 +1,213 @@
+/*
+ *	linux/arch/x86_64/kernel/irq.c
+ *
+ *	Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar
+ *
+ * This file contains the lowest level x86_64-specific interrupt
+ * entry and irq statistics code. All the remaining irq logic is
+ * done by the generic kernel/irq/ code and in the
+ * x86_64-specific irq controller code. (e.g. i8259.c and
+ * io_apic.c.)
+ */
+
+#include <linux/kernel_stat.h>
+#include <linux/interrupt.h>
+#include <linux/seq_file.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <asm/uaccess.h>
+#include <asm/io_apic.h>
+#include <asm/idle.h>
+#include <asm/smp.h>
+
+atomic_t irq_err_count;
+
+#ifdef CONFIG_DEBUG_STACKOVERFLOW
+/*
+ * Probabilistic stack overflow check:
+ *
+ * Only check the stack in process context, because everything else
+ * runs on the big interrupt stacks. Checking reliably is too expensive,
+ * so we just check from interrupts.
+ */
+static inline void stack_overflow_check(struct pt_regs *regs)
+{
+	u64 curbase = (u64)task_stack_page(current);
+	static unsigned long warned = -60*HZ;
+
+	if (regs->rsp >= curbase && regs->rsp <= curbase + THREAD_SIZE &&
+	    regs->rsp <  curbase + sizeof(struct thread_info) + 128 &&
+	    time_after(jiffies, warned + 60*HZ)) {
+		printk("do_IRQ: %s near stack overflow (cur:%Lx,rsp:%lx)\n",
+		       current->comm, curbase, regs->rsp);
+		show_stack(NULL,NULL);
+		warned = jiffies;
+	}
+}
+#endif
+
+/*
+ * Generic, controller-independent functions:
+ */
+
+int show_interrupts(struct seq_file *p, void *v)
+{
+	int i = *(loff_t *) v, j;
+	struct irqaction * action;
+	unsigned long flags;
+
+	if (i == 0) {
+		seq_printf(p, "           ");
+		for_each_online_cpu(j)
+			seq_printf(p, "CPU%-8d",j);
+		seq_putc(p, '\n');
+	}
+
+	if (i < NR_IRQS) {
+		spin_lock_irqsave(&irq_desc[i].lock, flags);
+		action = irq_desc[i].action;
+		if (!action) 
+			goto skip;
+		seq_printf(p, "%3d: ",i);
+#ifndef CONFIG_SMP
+		seq_printf(p, "%10u ", kstat_irqs(i));
+#else
+		for_each_online_cpu(j)
+			seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
+#endif
+		seq_printf(p, " %8s", 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');
+skip:
+		spin_unlock_irqrestore(&irq_desc[i].lock, flags);
+	} else if (i == NR_IRQS) {
+		seq_printf(p, "NMI: ");
+		for_each_online_cpu(j)
+			seq_printf(p, "%10u ", cpu_pda(j)->__nmi_count);
+		seq_putc(p, '\n');
+		seq_printf(p, "LOC: ");
+		for_each_online_cpu(j)
+			seq_printf(p, "%10u ", cpu_pda(j)->apic_timer_irqs);
+		seq_putc(p, '\n');
+		seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
+	}
+	return 0;
+}
+
+/*
+ * do_IRQ handles all normal device IRQ's (the special
+ * SMP cross-CPU interrupts have their own specific
+ * handlers).
+ */
+asmlinkage unsigned int do_IRQ(struct pt_regs *regs)
+{
+	struct pt_regs *old_regs = set_irq_regs(regs);
+
+	/* high bit used in ret_from_ code  */
+	unsigned vector = ~regs->orig_rax;
+	unsigned irq;
+
+	exit_idle();
+	irq_enter();
+	irq = __get_cpu_var(vector_irq)[vector];
+
+#ifdef CONFIG_DEBUG_STACKOVERFLOW
+	stack_overflow_check(regs);
+#endif
+
+	if (likely(irq < NR_IRQS))
+		generic_handle_irq(irq);
+	else {
+		if (!disable_apic)
+			ack_APIC_irq();
+
+		if (printk_ratelimit())
+			printk(KERN_EMERG "%s: %d.%d No irq handler for vector\n",
+				__func__, smp_processor_id(), vector);
+	}
+
+	irq_exit();
+
+	set_irq_regs(old_regs);
+	return 1;
+}
+
+#ifdef CONFIG_HOTPLUG_CPU
+void fixup_irqs(cpumask_t map)
+{
+	unsigned int irq;
+	static int warned;
+
+	for (irq = 0; irq < NR_IRQS; irq++) {
+		cpumask_t mask;
+		int break_affinity = 0;
+		int set_affinity = 1;
+
+		if (irq == 2)
+			continue;
+
+		/* interrupt's are disabled at this point */
+		spin_lock(&irq_desc[irq].lock);
+
+		if (!irq_has_action(irq) ||
+		    cpus_equal(irq_desc[irq].affinity, map)) {
+			spin_unlock(&irq_desc[irq].lock);
+			continue;
+		}
+
+		cpus_and(mask, irq_desc[irq].affinity, map);
+		if (cpus_empty(mask)) {
+			break_affinity = 1;
+			mask = map;
+		}
+
+		if (irq_desc[irq].chip->mask)
+			irq_desc[irq].chip->mask(irq);
+
+		if (irq_desc[irq].chip->set_affinity)
+			irq_desc[irq].chip->set_affinity(irq, mask);
+		else if (!(warned++))
+			set_affinity = 0;
+
+		if (irq_desc[irq].chip->unmask)
+			irq_desc[irq].chip->unmask(irq);
+
+		spin_unlock(&irq_desc[irq].lock);
+
+		if (break_affinity && set_affinity)
+			printk("Broke affinity for irq %i\n", irq);
+		else if (!set_affinity)
+			printk("Cannot set affinity for irq %i\n", irq);
+	}
+
+	/* That doesn't seem sufficient.  Give it 1ms. */
+	local_irq_enable();
+	mdelay(1);
+	local_irq_disable();
+}
+#endif
+
+extern void call_softirq(void);
+
+asmlinkage void do_softirq(void)
+{
+ 	__u32 pending;
+ 	unsigned long flags;
+
+ 	if (in_interrupt())
+ 		return;
+
+ 	local_irq_save(flags);
+ 	pending = local_softirq_pending();
+ 	/* Switch to interrupt stack */
+ 	if (pending) {
+		call_softirq();
+		WARN_ON_ONCE(softirq_count());
+	}
+ 	local_irq_restore(flags);
+}
+EXPORT_SYMBOL(do_softirq);

arch/x86/kernel/k8.c

@@ -0,0 +1,123 @@
+/*
+ * Shared support code for AMD K8 northbridges and derivates.
+ * Copyright 2006 Andi Kleen, SUSE Labs. Subject to GPLv2.
+ */
+#include <linux/gfp.h>
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <asm/k8.h>
+
+int num_k8_northbridges;
+EXPORT_SYMBOL(num_k8_northbridges);
+
+static u32 *flush_words;
+
+struct pci_device_id k8_nb_ids[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, 0x1103) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, 0x1203) },
+	{}
+};
+EXPORT_SYMBOL(k8_nb_ids);
+
+struct pci_dev **k8_northbridges;
+EXPORT_SYMBOL(k8_northbridges);
+
+static struct pci_dev *next_k8_northbridge(struct pci_dev *dev)
+{
+	do {
+		dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev);
+		if (!dev)
+			break;
+	} while (!pci_match_id(&k8_nb_ids[0], dev));
+	return dev;
+}
+
+int cache_k8_northbridges(void)
+{
+	int i;
+	struct pci_dev *dev;
+
+	if (num_k8_northbridges)
+		return 0;
+
+	dev = NULL;
+	while ((dev = next_k8_northbridge(dev)) != NULL)
+		num_k8_northbridges++;
+
+	k8_northbridges = kmalloc((num_k8_northbridges + 1) * sizeof(void *),
+				  GFP_KERNEL);
+	if (!k8_northbridges)
+		return -ENOMEM;
+
+	if (!num_k8_northbridges) {
+		k8_northbridges[0] = NULL;
+		return 0;
+	}
+
+	flush_words = kmalloc(num_k8_northbridges * sizeof(u32), GFP_KERNEL);
+	if (!flush_words) {
+		kfree(k8_northbridges);
+		return -ENOMEM;
+	}
+
+	dev = NULL;
+	i = 0;
+	while ((dev = next_k8_northbridge(dev)) != NULL) {
+		k8_northbridges[i] = dev;
+		pci_read_config_dword(dev, 0x9c, &flush_words[i++]);
+	}
+	k8_northbridges[i] = NULL;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(cache_k8_northbridges);
+
+/* Ignores subdevice/subvendor but as far as I can figure out
+   they're useless anyways */
+int __init early_is_k8_nb(u32 device)
+{
+	struct pci_device_id *id;
+	u32 vendor = device & 0xffff;
+	device >>= 16;
+	for (id = k8_nb_ids; id->vendor; id++)
+		if (vendor == id->vendor && device == id->device)
+			return 1;
+	return 0;
+}
+
+void k8_flush_garts(void)
+{
+	int flushed, i;
+	unsigned long flags;
+	static DEFINE_SPINLOCK(gart_lock);
+
+	/* Avoid races between AGP and IOMMU. In theory it's not needed
+	   but I'm not sure if the hardware won't lose flush requests
+	   when another is pending. This whole thing is so expensive anyways
+	   that it doesn't matter to serialize more. -AK */
+	spin_lock_irqsave(&gart_lock, flags);
+	flushed = 0;
+	for (i = 0; i < num_k8_northbridges; i++) {
+		pci_write_config_dword(k8_northbridges[i], 0x9c,
+				       flush_words[i]|1);
+		flushed++;
+	}
+	for (i = 0; i < num_k8_northbridges; i++) {
+		u32 w;
+		/* Make sure the hardware actually executed the flush*/
+		for (;;) {
+			pci_read_config_dword(k8_northbridges[i],
+					      0x9c, &w);
+			if (!(w & 1))
+				break;
+			cpu_relax();
+		}
+	}
+	spin_unlock_irqrestore(&gart_lock, flags);
+	if (!flushed)
+		printk("nothing to flush?\n");
+}
+EXPORT_SYMBOL_GPL(k8_flush_garts);
+

arch/x86/kernel/kprobes_64.c

@@ -0,0 +1,749 @@
+/*
+ *  Kernel Probes (KProbes)
+ *  arch/x86_64/kernel/kprobes.c
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright (C) IBM Corporation, 2002, 2004
+ *
+ * 2002-Oct	Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
+ *		Probes initial implementation ( includes contributions from
+ *		Rusty Russell).
+ * 2004-July	Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes
+ *		interface to access function arguments.
+ * 2004-Oct	Jim Keniston <kenistoj@us.ibm.com> and Prasanna S Panchamukhi
+ *		<prasanna@in.ibm.com> adapted for x86_64
+ * 2005-Mar	Roland McGrath <roland@redhat.com>
+ *		Fixed to handle %rip-relative addressing mode correctly.
+ * 2005-May     Rusty Lynch <rusty.lynch@intel.com>
+ *              Added function return probes functionality
+ */
+
+#include <linux/kprobes.h>
+#include <linux/ptrace.h>
+#include <linux/string.h>
+#include <linux/slab.h>
+#include <linux/preempt.h>
+#include <linux/module.h>
+#include <linux/kdebug.h>
+
+#include <asm/pgtable.h>
+#include <asm/uaccess.h>
+#include <asm/alternative.h>
+
+void jprobe_return_end(void);
+static void __kprobes arch_copy_kprobe(struct kprobe *p);
+
+DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
+DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
+
+/*
+ * returns non-zero if opcode modifies the interrupt flag.
+ */
+static __always_inline int is_IF_modifier(kprobe_opcode_t *insn)
+{
+	switch (*insn) {
+	case 0xfa:		/* cli */
+	case 0xfb:		/* sti */
+	case 0xcf:		/* iret/iretd */
+	case 0x9d:		/* popf/popfd */
+		return 1;
+	}
+
+	if (*insn  >= 0x40 && *insn <= 0x4f && *++insn == 0xcf)
+		return 1;
+	return 0;
+}
+
+int __kprobes arch_prepare_kprobe(struct kprobe *p)
+{
+	/* insn: must be on special executable page on x86_64. */
+	p->ainsn.insn = get_insn_slot();
+	if (!p->ainsn.insn) {
+		return -ENOMEM;
+	}
+	arch_copy_kprobe(p);
+	return 0;
+}
+
+/*
+ * Determine if the instruction uses the %rip-relative addressing mode.
+ * If it does, return the address of the 32-bit displacement word.
+ * If not, return null.
+ */
+static s32 __kprobes *is_riprel(u8 *insn)
+{
+#define W(row,b0,b1,b2,b3,b4,b5,b6,b7,b8,b9,ba,bb,bc,bd,be,bf)		      \
+	(((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) |   \
+	  (b4##UL << 0x4)|(b5##UL << 0x5)|(b6##UL << 0x6)|(b7##UL << 0x7) |   \
+	  (b8##UL << 0x8)|(b9##UL << 0x9)|(ba##UL << 0xa)|(bb##UL << 0xb) |   \
+	  (bc##UL << 0xc)|(bd##UL << 0xd)|(be##UL << 0xe)|(bf##UL << 0xf))    \
+	 << (row % 64))
+	static const u64 onebyte_has_modrm[256 / 64] = {
+		/*      0 1 2 3 4 5 6 7 8 9 a b c d e f         */
+		/*      -------------------------------         */
+		W(0x00, 1,1,1,1,0,0,0,0,1,1,1,1,0,0,0,0)| /* 00 */
+		W(0x10, 1,1,1,1,0,0,0,0,1,1,1,1,0,0,0,0)| /* 10 */
+		W(0x20, 1,1,1,1,0,0,0,0,1,1,1,1,0,0,0,0)| /* 20 */
+		W(0x30, 1,1,1,1,0,0,0,0,1,1,1,1,0,0,0,0), /* 30 */
+		W(0x40, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)| /* 40 */
+		W(0x50, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)| /* 50 */
+		W(0x60, 0,0,1,1,0,0,0,0,0,1,0,1,0,0,0,0)| /* 60 */
+		W(0x70, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0), /* 70 */
+		W(0x80, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* 80 */
+		W(0x90, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)| /* 90 */
+		W(0xa0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)| /* a0 */
+		W(0xb0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0), /* b0 */
+		W(0xc0, 1,1,0,0,1,1,1,1,0,0,0,0,0,0,0,0)| /* c0 */
+		W(0xd0, 1,1,1,1,0,0,0,0,1,1,1,1,1,1,1,1)| /* d0 */
+		W(0xe0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)| /* e0 */
+		W(0xf0, 0,0,0,0,0,0,1,1,0,0,0,0,0,0,1,1)  /* f0 */
+		/*      -------------------------------         */
+		/*      0 1 2 3 4 5 6 7 8 9 a b c d e f         */
+	};
+	static const u64 twobyte_has_modrm[256 / 64] = {
+		/*      0 1 2 3 4 5 6 7 8 9 a b c d e f         */
+		/*      -------------------------------         */
+		W(0x00, 1,1,1,1,0,0,0,0,0,0,0,0,0,1,0,1)| /* 0f */
+		W(0x10, 1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0)| /* 1f */
+		W(0x20, 1,1,1,1,1,0,1,0,1,1,1,1,1,1,1,1)| /* 2f */
+		W(0x30, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0), /* 3f */
+		W(0x40, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* 4f */
+		W(0x50, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* 5f */
+		W(0x60, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* 6f */
+		W(0x70, 1,1,1,1,1,1,1,0,0,0,0,0,1,1,1,1), /* 7f */
+		W(0x80, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)| /* 8f */
+		W(0x90, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* 9f */
+		W(0xa0, 0,0,0,1,1,1,1,1,0,0,0,1,1,1,1,1)| /* af */
+		W(0xb0, 1,1,1,1,1,1,1,1,0,0,1,1,1,1,1,1), /* bf */
+		W(0xc0, 1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0)| /* cf */
+		W(0xd0, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* df */
+		W(0xe0, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)| /* ef */
+		W(0xf0, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0)  /* ff */
+		/*      -------------------------------         */
+		/*      0 1 2 3 4 5 6 7 8 9 a b c d e f         */
+	};
+#undef	W
+	int need_modrm;
+
+	/* Skip legacy instruction prefixes.  */
+	while (1) {
+		switch (*insn) {
+		case 0x66:
+		case 0x67:
+		case 0x2e:
+		case 0x3e:
+		case 0x26:
+		case 0x64:
+		case 0x65:
+		case 0x36:
+		case 0xf0:
+		case 0xf3:
+		case 0xf2:
+			++insn;
+			continue;
+		}
+		break;
+	}
+
+	/* Skip REX instruction prefix.  */
+	if ((*insn & 0xf0) == 0x40)
+		++insn;
+
+	if (*insn == 0x0f) {	/* Two-byte opcode.  */
+		++insn;
+		need_modrm = test_bit(*insn, twobyte_has_modrm);
+	} else {		/* One-byte opcode.  */
+		need_modrm = test_bit(*insn, onebyte_has_modrm);
+	}
+
+	if (need_modrm) {
+		u8 modrm = *++insn;
+		if ((modrm & 0xc7) == 0x05) { /* %rip+disp32 addressing mode */
+			/* Displacement follows ModRM byte.  */
+			return (s32 *) ++insn;
+		}
+	}
+
+	/* No %rip-relative addressing mode here.  */
+	return NULL;
+}
+
+static void __kprobes arch_copy_kprobe(struct kprobe *p)
+{
+	s32 *ripdisp;
+	memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE);
+	ripdisp = is_riprel(p->ainsn.insn);
+	if (ripdisp) {
+		/*
+		 * The copied instruction uses the %rip-relative
+		 * addressing mode.  Adjust the displacement for the
+		 * difference between the original location of this
+		 * instruction and the location of the copy that will
+		 * actually be run.  The tricky bit here is making sure
+		 * that the sign extension happens correctly in this
+		 * calculation, since we need a signed 32-bit result to
+		 * be sign-extended to 64 bits when it's added to the
+		 * %rip value and yield the same 64-bit result that the
+		 * sign-extension of the original signed 32-bit
+		 * displacement would have given.
+		 */
+		s64 disp = (u8 *) p->addr + *ripdisp - (u8 *) p->ainsn.insn;
+		BUG_ON((s64) (s32) disp != disp); /* Sanity check.  */
+		*ripdisp = disp;
+	}
+	p->opcode = *p->addr;
+}
+
+void __kprobes arch_arm_kprobe(struct kprobe *p)
+{
+	text_poke(p->addr, ((unsigned char []){BREAKPOINT_INSTRUCTION}), 1);
+}
+
+void __kprobes arch_disarm_kprobe(struct kprobe *p)
+{
+	text_poke(p->addr, &p->opcode, 1);
+}
+
+void __kprobes arch_remove_kprobe(struct kprobe *p)
+{
+	mutex_lock(&kprobe_mutex);
+	free_insn_slot(p->ainsn.insn, 0);
+	mutex_unlock(&kprobe_mutex);
+}
+
+static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+	kcb->prev_kprobe.kp = kprobe_running();
+	kcb->prev_kprobe.status = kcb->kprobe_status;
+	kcb->prev_kprobe.old_rflags = kcb->kprobe_old_rflags;
+	kcb->prev_kprobe.saved_rflags = kcb->kprobe_saved_rflags;
+}
+
+static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+	__get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
+	kcb->kprobe_status = kcb->prev_kprobe.status;
+	kcb->kprobe_old_rflags = kcb->prev_kprobe.old_rflags;
+	kcb->kprobe_saved_rflags = kcb->prev_kprobe.saved_rflags;
+}
+
+static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
+				struct kprobe_ctlblk *kcb)
+{
+	__get_cpu_var(current_kprobe) = p;
+	kcb->kprobe_saved_rflags = kcb->kprobe_old_rflags
+		= (regs->eflags & (TF_MASK | IF_MASK));
+	if (is_IF_modifier(p->ainsn.insn))
+		kcb->kprobe_saved_rflags &= ~IF_MASK;
+}
+
+static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
+{
+	regs->eflags |= TF_MASK;
+	regs->eflags &= ~IF_MASK;
+	/*single step inline if the instruction is an int3*/
+	if (p->opcode == BREAKPOINT_INSTRUCTION)
+		regs->rip = (unsigned long)p->addr;
+	else
+		regs->rip = (unsigned long)p->ainsn.insn;
+}
+
+/* Called with kretprobe_lock held */
+void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
+				      struct pt_regs *regs)
+{
+	unsigned long *sara = (unsigned long *)regs->rsp;
+
+	ri->ret_addr = (kprobe_opcode_t *) *sara;
+	/* Replace the return addr with trampoline addr */
+	*sara = (unsigned long) &kretprobe_trampoline;
+}
+
+int __kprobes kprobe_handler(struct pt_regs *regs)
+{
+	struct kprobe *p;
+	int ret = 0;
+	kprobe_opcode_t *addr = (kprobe_opcode_t *)(regs->rip - sizeof(kprobe_opcode_t));
+	struct kprobe_ctlblk *kcb;
+
+	/*
+	 * We don't want to be preempted for the entire
+	 * duration of kprobe processing
+	 */
+	preempt_disable();
+	kcb = get_kprobe_ctlblk();
+
+	/* Check we're not actually recursing */
+	if (kprobe_running()) {
+		p = get_kprobe(addr);
+		if (p) {
+			if (kcb->kprobe_status == KPROBE_HIT_SS &&
+				*p->ainsn.insn == BREAKPOINT_INSTRUCTION) {
+				regs->eflags &= ~TF_MASK;
+				regs->eflags |= kcb->kprobe_saved_rflags;
+				goto no_kprobe;
+			} else if (kcb->kprobe_status == KPROBE_HIT_SSDONE) {
+				/* TODO: Provide re-entrancy from
+				 * post_kprobes_handler() and avoid exception
+				 * stack corruption while single-stepping on
+				 * the instruction of the new probe.
+				 */
+				arch_disarm_kprobe(p);
+				regs->rip = (unsigned long)p->addr;
+				reset_current_kprobe();
+				ret = 1;
+			} else {
+				/* We have reentered the kprobe_handler(), since
+				 * another probe was hit while within the
+				 * handler. We here save the original kprobe
+				 * variables and just single step on instruction
+				 * of the new probe without calling any user
+				 * handlers.
+				 */
+				save_previous_kprobe(kcb);
+				set_current_kprobe(p, regs, kcb);
+				kprobes_inc_nmissed_count(p);
+				prepare_singlestep(p, regs);
+				kcb->kprobe_status = KPROBE_REENTER;
+				return 1;
+			}
+		} else {
+			if (*addr != BREAKPOINT_INSTRUCTION) {
+			/* The breakpoint instruction was removed by
+			 * another cpu right after we hit, no further
+			 * handling of this interrupt is appropriate
+			 */
+				regs->rip = (unsigned long)addr;
+				ret = 1;
+				goto no_kprobe;
+			}
+			p = __get_cpu_var(current_kprobe);
+			if (p->break_handler && p->break_handler(p, regs)) {
+				goto ss_probe;
+			}
+		}
+		goto no_kprobe;
+	}
+
+	p = get_kprobe(addr);
+	if (!p) {
+		if (*addr != BREAKPOINT_INSTRUCTION) {
+			/*
+			 * The breakpoint instruction was removed right
+			 * after we hit it.  Another cpu has removed
+			 * either a probepoint or a debugger breakpoint
+			 * at this address.  In either case, no further
+			 * handling of this interrupt is appropriate.
+			 * Back up over the (now missing) int3 and run
+			 * the original instruction.
+			 */
+			regs->rip = (unsigned long)addr;
+			ret = 1;
+		}
+		/* Not one of ours: let kernel handle it */
+		goto no_kprobe;
+	}
+
+	set_current_kprobe(p, regs, kcb);
+	kcb->kprobe_status = KPROBE_HIT_ACTIVE;
+
+	if (p->pre_handler && p->pre_handler(p, regs))
+		/* handler has already set things up, so skip ss setup */
+		return 1;
+
+ss_probe:
+	prepare_singlestep(p, regs);
+	kcb->kprobe_status = KPROBE_HIT_SS;
+	return 1;
+
+no_kprobe:
+	preempt_enable_no_resched();
+	return ret;
+}
+
+/*
+ * For function-return probes, init_kprobes() establishes a probepoint
+ * here. When a retprobed function returns, this probe is hit and
+ * trampoline_probe_handler() runs, calling the kretprobe's handler.
+ */
+ void kretprobe_trampoline_holder(void)
+ {
+ 	asm volatile (  ".global kretprobe_trampoline\n"
+ 			"kretprobe_trampoline: \n"
+ 			"nop\n");
+ }
+
+/*
+ * Called when we hit the probe point at kretprobe_trampoline
+ */
+int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
+{
+	struct kretprobe_instance *ri = NULL;
+	struct hlist_head *head, empty_rp;
+	struct hlist_node *node, *tmp;
+	unsigned long flags, orig_ret_address = 0;
+	unsigned long trampoline_address =(unsigned long)&kretprobe_trampoline;
+
+	INIT_HLIST_HEAD(&empty_rp);
+	spin_lock_irqsave(&kretprobe_lock, flags);
+	head = kretprobe_inst_table_head(current);
+
+	/*
+	 * It is possible to have multiple instances associated with a given
+	 * task either because an multiple functions in the call path
+	 * have a return probe installed on them, and/or more then one return
+	 * return probe was registered for a target function.
+	 *
+	 * We can handle this because:
+	 *     - instances are always inserted at the head of the list
+	 *     - when multiple return probes are registered for the same
+	 *       function, the first instance's ret_addr will point to the
+	 *       real return address, and all the rest will point to
+	 *       kretprobe_trampoline
+	 */
+	hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
+		if (ri->task != current)
+			/* another task is sharing our hash bucket */
+			continue;
+
+		if (ri->rp && ri->rp->handler)
+			ri->rp->handler(ri, regs);
+
+		orig_ret_address = (unsigned long)ri->ret_addr;
+		recycle_rp_inst(ri, &empty_rp);
+
+		if (orig_ret_address != trampoline_address)
+			/*
+			 * This is the real return address. Any other
+			 * instances associated with this task are for
+			 * other calls deeper on the call stack
+			 */
+			break;
+	}
+
+	kretprobe_assert(ri, orig_ret_address, trampoline_address);
+	regs->rip = orig_ret_address;
+
+	reset_current_kprobe();
+	spin_unlock_irqrestore(&kretprobe_lock, flags);
+	preempt_enable_no_resched();
+
+	hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) {
+		hlist_del(&ri->hlist);
+		kfree(ri);
+	}
+	/*
+	 * By returning a non-zero value, we are telling
+	 * kprobe_handler() that we don't want the post_handler
+	 * to run (and have re-enabled preemption)
+	 */
+	return 1;
+}
+
+/*
+ * Called after single-stepping.  p->addr is the address of the
+ * instruction whose first byte has been replaced by the "int 3"
+ * instruction.  To avoid the SMP problems that can occur when we
+ * temporarily put back the original opcode to single-step, we
+ * single-stepped a copy of the instruction.  The address of this
+ * copy is p->ainsn.insn.
+ *
+ * This function prepares to return from the post-single-step
+ * interrupt.  We have to fix up the stack as follows:
+ *
+ * 0) Except in the case of absolute or indirect jump or call instructions,
+ * the new rip is relative to the copied instruction.  We need to make
+ * it relative to the original instruction.
+ *
+ * 1) If the single-stepped instruction was pushfl, then the TF and IF
+ * flags are set in the just-pushed eflags, and may need to be cleared.
+ *
+ * 2) If the single-stepped instruction was a call, the return address
+ * that is atop the stack is the address following the copied instruction.
+ * We need to make it the address following the original instruction.
+ */
+static void __kprobes resume_execution(struct kprobe *p,
+		struct pt_regs *regs, struct kprobe_ctlblk *kcb)
+{
+	unsigned long *tos = (unsigned long *)regs->rsp;
+	unsigned long next_rip = 0;
+	unsigned long copy_rip = (unsigned long)p->ainsn.insn;
+	unsigned long orig_rip = (unsigned long)p->addr;
+	kprobe_opcode_t *insn = p->ainsn.insn;
+
+	/*skip the REX prefix*/
+	if (*insn >= 0x40 && *insn <= 0x4f)
+		insn++;
+
+	switch (*insn) {
+	case 0x9c:		/* pushfl */
+		*tos &= ~(TF_MASK | IF_MASK);
+		*tos |= kcb->kprobe_old_rflags;
+		break;
+	case 0xc3:		/* ret/lret */
+	case 0xcb:
+	case 0xc2:
+	case 0xca:
+		regs->eflags &= ~TF_MASK;
+		/* rip is already adjusted, no more changes required*/
+		return;
+	case 0xe8:		/* call relative - Fix return addr */
+		*tos = orig_rip + (*tos - copy_rip);
+		break;
+	case 0xff:
+		if ((insn[1] & 0x30) == 0x10) {
+			/* call absolute, indirect */
+			/* Fix return addr; rip is correct. */
+			next_rip = regs->rip;
+			*tos = orig_rip + (*tos - copy_rip);
+		} else if (((insn[1] & 0x31) == 0x20) ||	/* jmp near, absolute indirect */
+			   ((insn[1] & 0x31) == 0x21)) {	/* jmp far, absolute indirect */
+			/* rip is correct. */
+			next_rip = regs->rip;
+		}
+		break;
+	case 0xea:		/* jmp absolute -- rip is correct */
+		next_rip = regs->rip;
+		break;
+	default:
+		break;
+	}
+
+	regs->eflags &= ~TF_MASK;
+	if (next_rip) {
+		regs->rip = next_rip;
+	} else {
+		regs->rip = orig_rip + (regs->rip - copy_rip);
+	}
+}
+
+int __kprobes post_kprobe_handler(struct pt_regs *regs)
+{
+	struct kprobe *cur = kprobe_running();
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+	if (!cur)
+		return 0;
+
+	if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) {
+		kcb->kprobe_status = KPROBE_HIT_SSDONE;
+		cur->post_handler(cur, regs, 0);
+	}
+
+	resume_execution(cur, regs, kcb);
+	regs->eflags |= kcb->kprobe_saved_rflags;
+
+	/* Restore the original saved kprobes variables and continue. */
+	if (kcb->kprobe_status == KPROBE_REENTER) {
+		restore_previous_kprobe(kcb);
+		goto out;
+	}
+	reset_current_kprobe();
+out:
+	preempt_enable_no_resched();
+
+	/*
+	 * if somebody else is singlestepping across a probe point, eflags
+	 * will have TF set, in which case, continue the remaining processing
+	 * of do_debug, as if this is not a probe hit.
+	 */
+	if (regs->eflags & TF_MASK)
+		return 0;
+
+	return 1;
+}
+
+int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
+{
+	struct kprobe *cur = kprobe_running();
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+	const struct exception_table_entry *fixup;
+
+	switch(kcb->kprobe_status) {
+	case KPROBE_HIT_SS:
+	case KPROBE_REENTER:
+		/*
+		 * We are here because the instruction being single
+		 * stepped caused a page fault. We reset the current
+		 * kprobe and the rip points back to the probe address
+		 * and allow the page fault handler to continue as a
+		 * normal page fault.
+		 */
+		regs->rip = (unsigned long)cur->addr;
+		regs->eflags |= kcb->kprobe_old_rflags;
+		if (kcb->kprobe_status == KPROBE_REENTER)
+			restore_previous_kprobe(kcb);
+		else
+			reset_current_kprobe();
+		preempt_enable_no_resched();
+		break;
+	case KPROBE_HIT_ACTIVE:
+	case KPROBE_HIT_SSDONE:
+		/*
+		 * We increment the nmissed count for accounting,
+		 * we can also use npre/npostfault count for accouting
+		 * these specific fault cases.
+		 */
+		kprobes_inc_nmissed_count(cur);
+
+		/*
+		 * We come here because instructions in the pre/post
+		 * handler caused the page_fault, this could happen
+		 * if handler tries to access user space by
+		 * copy_from_user(), get_user() etc. Let the
+		 * user-specified handler try to fix it first.
+		 */
+		if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr))
+			return 1;
+
+		/*
+		 * In case the user-specified fault handler returned
+		 * zero, try to fix up.
+		 */
+		fixup = search_exception_tables(regs->rip);
+		if (fixup) {
+			regs->rip = fixup->fixup;
+			return 1;
+		}
+
+		/*
+		 * fixup() could not handle it,
+		 * Let do_page_fault() fix it.
+		 */
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+/*
+ * Wrapper routine for handling exceptions.
+ */
+int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
+				       unsigned long val, void *data)
+{
+	struct die_args *args = (struct die_args *)data;
+	int ret = NOTIFY_DONE;
+
+	if (args->regs && user_mode(args->regs))
+		return ret;
+
+	switch (val) {
+	case DIE_INT3:
+		if (kprobe_handler(args->regs))
+			ret = NOTIFY_STOP;
+		break;
+	case DIE_DEBUG:
+		if (post_kprobe_handler(args->regs))
+			ret = NOTIFY_STOP;
+		break;
+	case DIE_GPF:
+	case DIE_PAGE_FAULT:
+		/* kprobe_running() needs smp_processor_id() */
+		preempt_disable();
+		if (kprobe_running() &&
+		    kprobe_fault_handler(args->regs, args->trapnr))
+			ret = NOTIFY_STOP;
+		preempt_enable();
+		break;
+	default:
+		break;
+	}
+	return ret;
+}
+
+int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
+{
+	struct jprobe *jp = container_of(p, struct jprobe, kp);
+	unsigned long addr;
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+	kcb->jprobe_saved_regs = *regs;
+	kcb->jprobe_saved_rsp = (long *) regs->rsp;
+	addr = (unsigned long)(kcb->jprobe_saved_rsp);
+	/*
+	 * As Linus pointed out, gcc assumes that the callee
+	 * owns the argument space and could overwrite it, e.g.
+	 * tailcall optimization. So, to be absolutely safe
+	 * we also save and restore enough stack bytes to cover
+	 * the argument area.
+	 */
+	memcpy(kcb->jprobes_stack, (kprobe_opcode_t *)addr,
+			MIN_STACK_SIZE(addr));
+	regs->eflags &= ~IF_MASK;
+	regs->rip = (unsigned long)(jp->entry);
+	return 1;
+}
+
+void __kprobes jprobe_return(void)
+{
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+	asm volatile ("       xchg   %%rbx,%%rsp     \n"
+		      "       int3			\n"
+		      "       .globl jprobe_return_end	\n"
+		      "       jprobe_return_end:	\n"
+		      "       nop			\n"::"b"
+		      (kcb->jprobe_saved_rsp):"memory");
+}
+
+int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
+{
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+	u8 *addr = (u8 *) (regs->rip - 1);
+	unsigned long stack_addr = (unsigned long)(kcb->jprobe_saved_rsp);
+	struct jprobe *jp = container_of(p, struct jprobe, kp);
+
+	if ((addr > (u8 *) jprobe_return) && (addr < (u8 *) jprobe_return_end)) {
+		if ((long *)regs->rsp != kcb->jprobe_saved_rsp) {
+			struct pt_regs *saved_regs =
+			    container_of(kcb->jprobe_saved_rsp,
+					    struct pt_regs, rsp);
+			printk("current rsp %p does not match saved rsp %p\n",
+			       (long *)regs->rsp, kcb->jprobe_saved_rsp);
+			printk("Saved registers for jprobe %p\n", jp);
+			show_registers(saved_regs);
+			printk("Current registers\n");
+			show_registers(regs);
+			BUG();
+		}
+		*regs = kcb->jprobe_saved_regs;
+		memcpy((kprobe_opcode_t *) stack_addr, kcb->jprobes_stack,
+		       MIN_STACK_SIZE(stack_addr));
+		preempt_enable_no_resched();
+		return 1;
+	}
+	return 0;
+}
+
+static struct kprobe trampoline_p = {
+	.addr = (kprobe_opcode_t *) &kretprobe_trampoline,
+	.pre_handler = trampoline_probe_handler
+};
+
+int __init arch_init_kprobes(void)
+{
+	return register_kprobe(&trampoline_p);
+}
+
+int __kprobes arch_trampoline_kprobe(struct kprobe *p)
+{
+	if (p->addr == (kprobe_opcode_t *)&kretprobe_trampoline)
+		return 1;
+
+	return 0;
+}

arch/x86/kernel/ldt_64.c

@@ -0,0 +1,252 @@
+/*
+ * linux/arch/x86_64/kernel/ldt.c
+ *
+ * Copyright (C) 1992 Krishna Balasubramanian and Linus Torvalds
+ * Copyright (C) 1999 Ingo Molnar <mingo@redhat.com>
+ * Copyright (C) 2002 Andi Kleen
+ * 
+ * This handles calls from both 32bit and 64bit mode.
+ */
+
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/string.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/vmalloc.h>
+#include <linux/slab.h>
+
+#include <asm/uaccess.h>
+#include <asm/system.h>
+#include <asm/ldt.h>
+#include <asm/desc.h>
+#include <asm/proto.h>
+
+#ifdef CONFIG_SMP /* avoids "defined but not used" warnig */
+static void flush_ldt(void *null)
+{
+	if (current->active_mm)
+               load_LDT(&current->active_mm->context);
+}
+#endif
+
+static int alloc_ldt(mm_context_t *pc, unsigned mincount, int reload)
+{
+	void *oldldt;
+	void *newldt;
+	unsigned oldsize;
+
+	if (mincount <= (unsigned)pc->size)
+		return 0;
+	oldsize = pc->size;
+	mincount = (mincount+511)&(~511);
+	if (mincount*LDT_ENTRY_SIZE > PAGE_SIZE)
+		newldt = vmalloc(mincount*LDT_ENTRY_SIZE);
+	else
+		newldt = kmalloc(mincount*LDT_ENTRY_SIZE, GFP_KERNEL);
+
+	if (!newldt)
+		return -ENOMEM;
+
+	if (oldsize)
+		memcpy(newldt, pc->ldt, oldsize*LDT_ENTRY_SIZE);
+	oldldt = pc->ldt;
+	memset(newldt+oldsize*LDT_ENTRY_SIZE, 0, (mincount-oldsize)*LDT_ENTRY_SIZE);
+	wmb();
+	pc->ldt = newldt;
+	wmb();
+	pc->size = mincount;
+	wmb();
+	if (reload) {
+#ifdef CONFIG_SMP
+		cpumask_t mask;
+
+		preempt_disable();
+		mask = cpumask_of_cpu(smp_processor_id());
+		load_LDT(pc);
+		if (!cpus_equal(current->mm->cpu_vm_mask, mask))
+			smp_call_function(flush_ldt, NULL, 1, 1);
+		preempt_enable();
+#else
+		load_LDT(pc);
+#endif
+	}
+	if (oldsize) {
+		if (oldsize*LDT_ENTRY_SIZE > PAGE_SIZE)
+			vfree(oldldt);
+		else
+			kfree(oldldt);
+	}
+	return 0;
+}
+
+static inline int copy_ldt(mm_context_t *new, mm_context_t *old)
+{
+	int err = alloc_ldt(new, old->size, 0);
+	if (err < 0)
+		return err;
+	memcpy(new->ldt, old->ldt, old->size*LDT_ENTRY_SIZE);
+	return 0;
+}
+
+/*
+ * we do not have to muck with descriptors here, that is
+ * done in switch_mm() as needed.
+ */
+int init_new_context(struct task_struct *tsk, struct mm_struct *mm)
+{
+	struct mm_struct * old_mm;
+	int retval = 0;
+
+	init_MUTEX(&mm->context.sem);
+	mm->context.size = 0;
+	old_mm = current->mm;
+	if (old_mm && old_mm->context.size > 0) {
+		down(&old_mm->context.sem);
+		retval = copy_ldt(&mm->context, &old_mm->context);
+		up(&old_mm->context.sem);
+	}
+	return retval;
+}
+
+/*
+ * 
+ * Don't touch the LDT register - we're already in the next thread.
+ */
+void destroy_context(struct mm_struct *mm)
+{
+	if (mm->context.size) {
+		if ((unsigned)mm->context.size*LDT_ENTRY_SIZE > PAGE_SIZE)
+			vfree(mm->context.ldt);
+		else
+			kfree(mm->context.ldt);
+		mm->context.size = 0;
+	}
+}
+
+static int read_ldt(void __user * ptr, unsigned long bytecount)
+{
+	int err;
+	unsigned long size;
+	struct mm_struct * mm = current->mm;
+
+	if (!mm->context.size)
+		return 0;
+	if (bytecount > LDT_ENTRY_SIZE*LDT_ENTRIES)
+		bytecount = LDT_ENTRY_SIZE*LDT_ENTRIES;
+
+	down(&mm->context.sem);
+	size = mm->context.size*LDT_ENTRY_SIZE;
+	if (size > bytecount)
+		size = bytecount;
+
+	err = 0;
+	if (copy_to_user(ptr, mm->context.ldt, size))
+		err = -EFAULT;
+	up(&mm->context.sem);
+	if (err < 0)
+		goto error_return;
+	if (size != bytecount) {
+		/* zero-fill the rest */
+		if (clear_user(ptr+size, bytecount-size) != 0) {
+			err = -EFAULT;
+			goto error_return;
+		}
+	}
+	return bytecount;
+error_return:
+	return err;
+}
+
+static int read_default_ldt(void __user * ptr, unsigned long bytecount)
+{
+	/* Arbitrary number */ 
+	/* x86-64 default LDT is all zeros */
+	if (bytecount > 128) 
+		bytecount = 128; 	
+	if (clear_user(ptr, bytecount))
+		return -EFAULT;
+	return bytecount; 
+}
+
+static int write_ldt(void __user * ptr, unsigned long bytecount, int oldmode)
+{
+	struct task_struct *me = current;
+	struct mm_struct * mm = me->mm;
+	__u32 entry_1, entry_2, *lp;
+	int error;
+	struct user_desc ldt_info;
+
+	error = -EINVAL;
+
+	if (bytecount != sizeof(ldt_info))
+		goto out;
+	error = -EFAULT; 	
+	if (copy_from_user(&ldt_info, ptr, bytecount))
+		goto out;
+
+	error = -EINVAL;
+	if (ldt_info.entry_number >= LDT_ENTRIES)
+		goto out;
+	if (ldt_info.contents == 3) {
+		if (oldmode)
+			goto out;
+		if (ldt_info.seg_not_present == 0)
+			goto out;
+	}
+
+	down(&mm->context.sem);
+	if (ldt_info.entry_number >= (unsigned)mm->context.size) {
+		error = alloc_ldt(&current->mm->context, ldt_info.entry_number+1, 1);
+		if (error < 0)
+			goto out_unlock;
+	}
+
+	lp = (__u32 *) ((ldt_info.entry_number << 3) + (char *) mm->context.ldt);
+
+   	/* Allow LDTs to be cleared by the user. */
+   	if (ldt_info.base_addr == 0 && ldt_info.limit == 0) {
+		if (oldmode || LDT_empty(&ldt_info)) {
+			entry_1 = 0;
+			entry_2 = 0;
+			goto install;
+		}
+	}
+
+	entry_1 = LDT_entry_a(&ldt_info);
+	entry_2 = LDT_entry_b(&ldt_info);
+	if (oldmode)
+		entry_2 &= ~(1 << 20);
+
+	/* Install the new entry ...  */
+install:
+	*lp	= entry_1;
+	*(lp+1)	= entry_2;
+	error = 0;
+
+out_unlock:
+	up(&mm->context.sem);
+out:
+	return error;
+}
+
+asmlinkage int sys_modify_ldt(int func, void __user *ptr, unsigned long bytecount)
+{
+	int ret = -ENOSYS;
+
+	switch (func) {
+	case 0:
+		ret = read_ldt(ptr, bytecount);
+		break;
+	case 1:
+		ret = write_ldt(ptr, bytecount, 1);
+		break;
+	case 2:
+		ret = read_default_ldt(ptr, bytecount);
+		break;
+	case 0x11:
+		ret = write_ldt(ptr, bytecount, 0);
+		break;
+	}
+	return ret;
+}

arch/x86/kernel/machine_kexec_64.c

@@ -0,0 +1,259 @@
+/*
+ * machine_kexec.c - handle transition of Linux booting another kernel
+ * Copyright (C) 2002-2005 Eric Biederman  <ebiederm@xmission.com>
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2.  See the file COPYING for more details.
+ */
+
+#include <linux/mm.h>
+#include <linux/kexec.h>
+#include <linux/string.h>
+#include <linux/reboot.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+#include <asm/mmu_context.h>
+#include <asm/io.h>
+
+#define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
+static u64 kexec_pgd[512] PAGE_ALIGNED;
+static u64 kexec_pud0[512] PAGE_ALIGNED;
+static u64 kexec_pmd0[512] PAGE_ALIGNED;
+static u64 kexec_pte0[512] PAGE_ALIGNED;
+static u64 kexec_pud1[512] PAGE_ALIGNED;
+static u64 kexec_pmd1[512] PAGE_ALIGNED;
+static u64 kexec_pte1[512] PAGE_ALIGNED;
+
+static void init_level2_page(pmd_t *level2p, unsigned long addr)
+{
+	unsigned long end_addr;
+
+	addr &= PAGE_MASK;
+	end_addr = addr + PUD_SIZE;
+	while (addr < end_addr) {
+		set_pmd(level2p++, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC));
+		addr += PMD_SIZE;
+	}
+}
+
+static int init_level3_page(struct kimage *image, pud_t *level3p,
+				unsigned long addr, unsigned long last_addr)
+{
+	unsigned long end_addr;
+	int result;
+
+	result = 0;
+	addr &= PAGE_MASK;
+	end_addr = addr + PGDIR_SIZE;
+	while ((addr < last_addr) && (addr < end_addr)) {
+		struct page *page;
+		pmd_t *level2p;
+
+		page = kimage_alloc_control_pages(image, 0);
+		if (!page) {
+			result = -ENOMEM;
+			goto out;
+		}
+		level2p = (pmd_t *)page_address(page);
+		init_level2_page(level2p, addr);
+		set_pud(level3p++, __pud(__pa(level2p) | _KERNPG_TABLE));
+		addr += PUD_SIZE;
+	}
+	/* clear the unused entries */
+	while (addr < end_addr) {
+		pud_clear(level3p++);
+		addr += PUD_SIZE;
+	}
+out:
+	return result;
+}
+
+
+static int init_level4_page(struct kimage *image, pgd_t *level4p,
+				unsigned long addr, unsigned long last_addr)
+{
+	unsigned long end_addr;
+	int result;
+
+	result = 0;
+	addr &= PAGE_MASK;
+	end_addr = addr + (PTRS_PER_PGD * PGDIR_SIZE);
+	while ((addr < last_addr) && (addr < end_addr)) {
+		struct page *page;
+		pud_t *level3p;
+
+		page = kimage_alloc_control_pages(image, 0);
+		if (!page) {
+			result = -ENOMEM;
+			goto out;
+		}
+		level3p = (pud_t *)page_address(page);
+		result = init_level3_page(image, level3p, addr, last_addr);
+		if (result) {
+			goto out;
+		}
+		set_pgd(level4p++, __pgd(__pa(level3p) | _KERNPG_TABLE));
+		addr += PGDIR_SIZE;
+	}
+	/* clear the unused entries */
+	while (addr < end_addr) {
+		pgd_clear(level4p++);
+		addr += PGDIR_SIZE;
+	}
+out:
+	return result;
+}
+
+
+static int init_pgtable(struct kimage *image, unsigned long start_pgtable)
+{
+	pgd_t *level4p;
+	level4p = (pgd_t *)__va(start_pgtable);
+ 	return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT);
+}
+
+static void set_idt(void *newidt, u16 limit)
+{
+	struct desc_ptr curidt;
+
+	/* x86-64 supports unaliged loads & stores */
+	curidt.size    = limit;
+	curidt.address = (unsigned long)newidt;
+
+	__asm__ __volatile__ (
+		"lidtq %0\n"
+		: : "m" (curidt)
+		);
+};
+
+
+static void set_gdt(void *newgdt, u16 limit)
+{
+	struct desc_ptr curgdt;
+
+	/* x86-64 supports unaligned loads & stores */
+	curgdt.size    = limit;
+	curgdt.address = (unsigned long)newgdt;
+
+	__asm__ __volatile__ (
+		"lgdtq %0\n"
+		: : "m" (curgdt)
+		);
+};
+
+static void load_segments(void)
+{
+	__asm__ __volatile__ (
+		"\tmovl %0,%%ds\n"
+		"\tmovl %0,%%es\n"
+		"\tmovl %0,%%ss\n"
+		"\tmovl %0,%%fs\n"
+		"\tmovl %0,%%gs\n"
+		: : "a" (__KERNEL_DS) : "memory"
+		);
+}
+
+int machine_kexec_prepare(struct kimage *image)
+{
+	unsigned long start_pgtable;
+	int result;
+
+	/* Calculate the offsets */
+	start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
+
+	/* Setup the identity mapped 64bit page table */
+	result = init_pgtable(image, start_pgtable);
+	if (result)
+		return result;
+
+	return 0;
+}
+
+void machine_kexec_cleanup(struct kimage *image)
+{
+	return;
+}
+
+/*
+ * Do not allocate memory (or fail in any way) in machine_kexec().
+ * We are past the point of no return, committed to rebooting now.
+ */
+NORET_TYPE void machine_kexec(struct kimage *image)
+{
+	unsigned long page_list[PAGES_NR];
+	void *control_page;
+
+	/* Interrupts aren't acceptable while we reboot */
+	local_irq_disable();
+
+	control_page = page_address(image->control_code_page) + PAGE_SIZE;
+	memcpy(control_page, relocate_kernel, PAGE_SIZE);
+
+	page_list[PA_CONTROL_PAGE] = virt_to_phys(control_page);
+	page_list[VA_CONTROL_PAGE] = (unsigned long)relocate_kernel;
+	page_list[PA_PGD] = virt_to_phys(&kexec_pgd);
+	page_list[VA_PGD] = (unsigned long)kexec_pgd;
+	page_list[PA_PUD_0] = virt_to_phys(&kexec_pud0);
+	page_list[VA_PUD_0] = (unsigned long)kexec_pud0;
+	page_list[PA_PMD_0] = virt_to_phys(&kexec_pmd0);
+	page_list[VA_PMD_0] = (unsigned long)kexec_pmd0;
+	page_list[PA_PTE_0] = virt_to_phys(&kexec_pte0);
+	page_list[VA_PTE_0] = (unsigned long)kexec_pte0;
+	page_list[PA_PUD_1] = virt_to_phys(&kexec_pud1);
+	page_list[VA_PUD_1] = (unsigned long)kexec_pud1;
+	page_list[PA_PMD_1] = virt_to_phys(&kexec_pmd1);
+	page_list[VA_PMD_1] = (unsigned long)kexec_pmd1;
+	page_list[PA_PTE_1] = virt_to_phys(&kexec_pte1);
+	page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
+
+	page_list[PA_TABLE_PAGE] =
+	  (unsigned long)__pa(page_address(image->control_code_page));
+
+	/* The segment registers are funny things, they have both a
+	 * visible and an invisible part.  Whenever the visible part is
+	 * set to a specific selector, the invisible part is loaded
+	 * with from a table in memory.  At no other time is the
+	 * descriptor table in memory accessed.
+	 *
+	 * I take advantage of this here by force loading the
+	 * segments, before I zap the gdt with an invalid value.
+	 */
+	load_segments();
+	/* The gdt & idt are now invalid.
+	 * If you want to load them you must set up your own idt & gdt.
+	 */
+	set_gdt(phys_to_virt(0),0);
+	set_idt(phys_to_virt(0),0);
+
+	/* now call it */
+	relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
+			image->start);
+}
+
+/* crashkernel=size@addr specifies the location to reserve for
+ * a crash kernel.  By reserving this memory we guarantee
+ * that linux never set's it up as a DMA target.
+ * Useful for holding code to do something appropriate
+ * after a kernel panic.
+ */
+static int __init setup_crashkernel(char *arg)
+{
+	unsigned long size, base;
+	char *p;
+	if (!arg)
+		return -EINVAL;
+	size = memparse(arg, &p);
+	if (arg == p)
+		return -EINVAL;
+	if (*p == '@') {
+		base = memparse(p+1, &p);
+		/* FIXME: Do I want a sanity check to validate the
+		 * memory range?  Yes you do, but it's too early for
+		 * e820 -AK */
+		crashk_res.start = base;
+		crashk_res.end   = base + size - 1;
+	}
+	return 0;
+}
+early_param("crashkernel", setup_crashkernel);
+

arch/x86/kernel/mce_64.c

@@ -0,0 +1,875 @@
+/*
+ * Machine check handler.
+ * K8 parts Copyright 2002,2003 Andi Kleen, SuSE Labs.
+ * Rest from unknown author(s). 
+ * 2004 Andi Kleen. Rewrote most of it. 
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/string.h>
+#include <linux/rcupdate.h>
+#include <linux/kallsyms.h>
+#include <linux/sysdev.h>
+#include <linux/miscdevice.h>
+#include <linux/fs.h>
+#include <linux/capability.h>
+#include <linux/cpu.h>
+#include <linux/percpu.h>
+#include <linux/poll.h>
+#include <linux/thread_info.h>
+#include <linux/ctype.h>
+#include <linux/kmod.h>
+#include <linux/kdebug.h>
+#include <asm/processor.h> 
+#include <asm/msr.h>
+#include <asm/mce.h>
+#include <asm/uaccess.h>
+#include <asm/smp.h>
+#include <asm/idle.h>
+
+#define MISC_MCELOG_MINOR 227
+#define NR_BANKS 6
+
+atomic_t mce_entry;
+
+static int mce_dont_init;
+
+/*
+ * Tolerant levels:
+ *   0: always panic on uncorrected errors, log corrected errors
+ *   1: panic or SIGBUS on uncorrected errors, log corrected errors
+ *   2: SIGBUS or log uncorrected errors (if possible), log corrected errors
+ *   3: never panic or SIGBUS, log all errors (for testing only)
+ */
+static int tolerant = 1;
+static int banks;
+static unsigned long bank[NR_BANKS] = { [0 ... NR_BANKS-1] = ~0UL };
+static unsigned long notify_user;
+static int rip_msr;
+static int mce_bootlog = 1;
+static atomic_t mce_events;
+
+static char trigger[128];
+static char *trigger_argv[2] = { trigger, NULL };
+
+static DECLARE_WAIT_QUEUE_HEAD(mce_wait);
+
+/*
+ * Lockless MCE logging infrastructure.
+ * This avoids deadlocks on printk locks without having to break locks. Also
+ * separate MCEs from kernel messages to avoid bogus bug reports.
+ */
+
+struct mce_log mcelog = { 
+	MCE_LOG_SIGNATURE,
+	MCE_LOG_LEN,
+}; 
+
+void mce_log(struct mce *mce)
+{
+	unsigned next, entry;
+	atomic_inc(&mce_events);
+	mce->finished = 0;
+	wmb();
+	for (;;) {
+		entry = rcu_dereference(mcelog.next);
+		/* The rmb forces the compiler to reload next in each
+		    iteration */
+		rmb();
+		for (;;) {
+			/* When the buffer fills up discard new entries. Assume
+			   that the earlier errors are the more interesting. */
+			if (entry >= MCE_LOG_LEN) {
+				set_bit(MCE_OVERFLOW, &mcelog.flags);
+				return;
+			}
+			/* Old left over entry. Skip. */
+			if (mcelog.entry[entry].finished) {
+				entry++;
+				continue;
+			}
+			break;
+		}
+		smp_rmb();
+		next = entry + 1;
+		if (cmpxchg(&mcelog.next, entry, next) == entry)
+			break;
+	}
+	memcpy(mcelog.entry + entry, mce, sizeof(struct mce));
+	wmb();
+	mcelog.entry[entry].finished = 1;
+	wmb();
+
+	set_bit(0, &notify_user);
+}
+
+static void print_mce(struct mce *m)
+{
+	printk(KERN_EMERG "\n"
+	       KERN_EMERG "HARDWARE ERROR\n"
+	       KERN_EMERG
+	       "CPU %d: Machine Check Exception: %16Lx Bank %d: %016Lx\n",
+	       m->cpu, m->mcgstatus, m->bank, m->status);
+	if (m->rip) {
+		printk(KERN_EMERG 
+		       "RIP%s %02x:<%016Lx> ",
+		       !(m->mcgstatus & MCG_STATUS_EIPV) ? " !INEXACT!" : "",
+		       m->cs, m->rip);
+		if (m->cs == __KERNEL_CS)
+			print_symbol("{%s}", m->rip);
+		printk("\n");
+	}
+	printk(KERN_EMERG "TSC %Lx ", m->tsc); 
+	if (m->addr)
+		printk("ADDR %Lx ", m->addr);
+	if (m->misc)
+		printk("MISC %Lx ", m->misc); 	
+	printk("\n");
+	printk(KERN_EMERG "This is not a software problem!\n");
+        printk(KERN_EMERG
+    "Run through mcelog --ascii to decode and contact your hardware vendor\n");
+}
+
+static void mce_panic(char *msg, struct mce *backup, unsigned long start)
+{ 
+	int i;
+
+	oops_begin();
+	for (i = 0; i < MCE_LOG_LEN; i++) {
+		unsigned long tsc = mcelog.entry[i].tsc;
+		if (time_before(tsc, start))
+			continue;
+		print_mce(&mcelog.entry[i]); 
+		if (backup && mcelog.entry[i].tsc == backup->tsc)
+			backup = NULL;
+	}
+	if (backup)
+		print_mce(backup);
+	panic(msg);
+} 
+
+static int mce_available(struct cpuinfo_x86 *c)
+{
+	return cpu_has(c, X86_FEATURE_MCE) && cpu_has(c, X86_FEATURE_MCA);
+}
+
+static inline void mce_get_rip(struct mce *m, struct pt_regs *regs)
+{
+	if (regs && (m->mcgstatus & MCG_STATUS_RIPV)) {
+		m->rip = regs->rip;
+		m->cs = regs->cs;
+	} else {
+		m->rip = 0;
+		m->cs = 0;
+	}
+	if (rip_msr) {
+		/* Assume the RIP in the MSR is exact. Is this true? */
+		m->mcgstatus |= MCG_STATUS_EIPV;
+		rdmsrl(rip_msr, m->rip);
+		m->cs = 0;
+	}
+}
+
+/* 
+ * The actual machine check handler
+ */
+
+void do_machine_check(struct pt_regs * regs, long error_code)
+{
+	struct mce m, panicm;
+	u64 mcestart = 0;
+	int i;
+	int panicm_found = 0;
+	/*
+	 * If no_way_out gets set, there is no safe way to recover from this
+	 * MCE.  If tolerant is cranked up, we'll try anyway.
+	 */
+	int no_way_out = 0;
+	/*
+	 * If kill_it gets set, there might be a way to recover from this
+	 * error.
+	 */
+	int kill_it = 0;
+
+	atomic_inc(&mce_entry);
+
+	if (regs)
+		notify_die(DIE_NMI, "machine check", regs, error_code, 18, SIGKILL);
+	if (!banks)
+		goto out2;
+
+	memset(&m, 0, sizeof(struct mce));
+	m.cpu = smp_processor_id();
+	rdmsrl(MSR_IA32_MCG_STATUS, m.mcgstatus);
+	/* if the restart IP is not valid, we're done for */
+	if (!(m.mcgstatus & MCG_STATUS_RIPV))
+		no_way_out = 1;
+	
+	rdtscll(mcestart);
+	barrier();
+
+	for (i = 0; i < banks; i++) {
+		if (!bank[i])
+			continue;
+		
+		m.misc = 0; 
+		m.addr = 0;
+		m.bank = i;
+		m.tsc = 0;
+
+		rdmsrl(MSR_IA32_MC0_STATUS + i*4, m.status);
+		if ((m.status & MCI_STATUS_VAL) == 0)
+			continue;
+
+		if (m.status & MCI_STATUS_EN) {
+			/* if PCC was set, there's no way out */
+			no_way_out |= !!(m.status & MCI_STATUS_PCC);
+			/*
+			 * If this error was uncorrectable and there was
+			 * an overflow, we're in trouble.  If no overflow,
+			 * we might get away with just killing a task.
+			 */
+			if (m.status & MCI_STATUS_UC) {
+				if (tolerant < 1 || m.status & MCI_STATUS_OVER)
+					no_way_out = 1;
+				kill_it = 1;
+			}
+		}
+
+		if (m.status & MCI_STATUS_MISCV)
+			rdmsrl(MSR_IA32_MC0_MISC + i*4, m.misc);
+		if (m.status & MCI_STATUS_ADDRV)
+			rdmsrl(MSR_IA32_MC0_ADDR + i*4, m.addr);
+
+		mce_get_rip(&m, regs);
+		if (error_code >= 0)
+			rdtscll(m.tsc);
+		if (error_code != -2)
+			mce_log(&m);
+
+		/* Did this bank cause the exception? */
+		/* Assume that the bank with uncorrectable errors did it,
+		   and that there is only a single one. */
+		if ((m.status & MCI_STATUS_UC) && (m.status & MCI_STATUS_EN)) {
+			panicm = m;
+			panicm_found = 1;
+		}
+
+		add_taint(TAINT_MACHINE_CHECK);
+	}
+
+	/* Never do anything final in the polling timer */
+	if (!regs)
+		goto out;
+
+	/* If we didn't find an uncorrectable error, pick
+	   the last one (shouldn't happen, just being safe). */
+	if (!panicm_found)
+		panicm = m;
+
+	/*
+	 * If we have decided that we just CAN'T continue, and the user
+	 *  has not set tolerant to an insane level, give up and die.
+	 */
+	if (no_way_out && tolerant < 3)
+		mce_panic("Machine check", &panicm, mcestart);
+
+	/*
+	 * If the error seems to be unrecoverable, something should be
+	 * done.  Try to kill as little as possible.  If we can kill just
+	 * one task, do that.  If the user has set the tolerance very
+	 * high, don't try to do anything at all.
+	 */
+	if (kill_it && tolerant < 3) {
+		int user_space = 0;
+
+		/*
+		 * If the EIPV bit is set, it means the saved IP is the
+		 * instruction which caused the MCE.
+		 */
+		if (m.mcgstatus & MCG_STATUS_EIPV)
+			user_space = panicm.rip && (panicm.cs & 3);
+
+		/*
+		 * If we know that the error was in user space, send a
+		 * SIGBUS.  Otherwise, panic if tolerance is low.
+		 *
+		 * do_exit() takes an awful lot of locks and has a slight
+		 * risk of deadlocking.
+		 */
+		if (user_space) {
+			do_exit(SIGBUS);
+		} else if (panic_on_oops || tolerant < 2) {
+			mce_panic("Uncorrected machine check",
+				&panicm, mcestart);
+		}
+	}
+
+	/* notify userspace ASAP */
+	set_thread_flag(TIF_MCE_NOTIFY);
+
+ out:
+	/* the last thing we do is clear state */
+	for (i = 0; i < banks; i++)
+		wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0);
+	wrmsrl(MSR_IA32_MCG_STATUS, 0);
+ out2:
+	atomic_dec(&mce_entry);
+}
+
+#ifdef CONFIG_X86_MCE_INTEL
+/***
+ * mce_log_therm_throt_event - Logs the thermal throttling event to mcelog
+ * @cpu: The CPU on which the event occured.
+ * @status: Event status information
+ *
+ * This function should be called by the thermal interrupt after the
+ * event has been processed and the decision was made to log the event
+ * further.
+ *
+ * The status parameter will be saved to the 'status' field of 'struct mce'
+ * and historically has been the register value of the
+ * MSR_IA32_THERMAL_STATUS (Intel) msr.
+ */
+void mce_log_therm_throt_event(unsigned int cpu, __u64 status)
+{
+	struct mce m;
+
+	memset(&m, 0, sizeof(m));
+	m.cpu = cpu;
+	m.bank = MCE_THERMAL_BANK;
+	m.status = status;
+	rdtscll(m.tsc);
+	mce_log(&m);
+}
+#endif /* CONFIG_X86_MCE_INTEL */
+
+/*
+ * Periodic polling timer for "silent" machine check errors.  If the
+ * poller finds an MCE, poll 2x faster.  When the poller finds no more
+ * errors, poll 2x slower (up to check_interval seconds).
+ */
+
+static int check_interval = 5 * 60; /* 5 minutes */
+static int next_interval; /* in jiffies */
+static void mcheck_timer(struct work_struct *work);
+static DECLARE_DELAYED_WORK(mcheck_work, mcheck_timer);
+
+static void mcheck_check_cpu(void *info)
+{
+	if (mce_available(&current_cpu_data))
+		do_machine_check(NULL, 0);
+}
+
+static void mcheck_timer(struct work_struct *work)
+{
+	on_each_cpu(mcheck_check_cpu, NULL, 1, 1);
+
+	/*
+	 * Alert userspace if needed.  If we logged an MCE, reduce the
+	 * polling interval, otherwise increase the polling interval.
+	 */
+	if (mce_notify_user()) {
+		next_interval = max(next_interval/2, HZ/100);
+	} else {
+		next_interval = min(next_interval*2,
+				(int)round_jiffies_relative(check_interval*HZ));
+	}
+
+	schedule_delayed_work(&mcheck_work, next_interval);
+}
+
+/*
+ * This is only called from process context.  This is where we do
+ * anything we need to alert userspace about new MCEs.  This is called
+ * directly from the poller and also from entry.S and idle, thanks to
+ * TIF_MCE_NOTIFY.
+ */
+int mce_notify_user(void)
+{
+	clear_thread_flag(TIF_MCE_NOTIFY);
+	if (test_and_clear_bit(0, &notify_user)) {
+		static unsigned long last_print;
+		unsigned long now = jiffies;
+
+		wake_up_interruptible(&mce_wait);
+		if (trigger[0])
+			call_usermodehelper(trigger, trigger_argv, NULL,
+						UMH_NO_WAIT);
+
+		if (time_after_eq(now, last_print + (check_interval*HZ))) {
+			last_print = now;
+			printk(KERN_INFO "Machine check events logged\n");
+		}
+
+		return 1;
+	}
+	return 0;
+}
+
+/* see if the idle task needs to notify userspace */
+static int
+mce_idle_callback(struct notifier_block *nfb, unsigned long action, void *junk)
+{
+	/* IDLE_END should be safe - interrupts are back on */
+	if (action == IDLE_END && test_thread_flag(TIF_MCE_NOTIFY))
+		mce_notify_user();
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block mce_idle_notifier = {
+	.notifier_call = mce_idle_callback,
+};
+
+static __init int periodic_mcheck_init(void)
+{ 
+	next_interval = check_interval * HZ;
+	if (next_interval)
+		schedule_delayed_work(&mcheck_work,
+				      round_jiffies_relative(next_interval));
+	idle_notifier_register(&mce_idle_notifier);
+	return 0;
+} 
+__initcall(periodic_mcheck_init);
+
+
+/* 
+ * Initialize Machine Checks for a CPU.
+ */
+static void mce_init(void *dummy)
+{
+	u64 cap;
+	int i;
+
+	rdmsrl(MSR_IA32_MCG_CAP, cap);
+	banks = cap & 0xff;
+	if (banks > NR_BANKS) { 
+		printk(KERN_INFO "MCE: warning: using only %d banks\n", banks);
+		banks = NR_BANKS; 
+	}
+	/* Use accurate RIP reporting if available. */
+	if ((cap & (1<<9)) && ((cap >> 16) & 0xff) >= 9)
+		rip_msr = MSR_IA32_MCG_EIP;
+
+	/* Log the machine checks left over from the previous reset.
+	   This also clears all registers */
+	do_machine_check(NULL, mce_bootlog ? -1 : -2);
+
+	set_in_cr4(X86_CR4_MCE);
+
+	if (cap & MCG_CTL_P)
+		wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff);
+
+	for (i = 0; i < banks; i++) {
+		wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]);
+		wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0);
+	}	
+}
+
+/* Add per CPU specific workarounds here */
+static void __cpuinit mce_cpu_quirks(struct cpuinfo_x86 *c)
+{ 
+	/* This should be disabled by the BIOS, but isn't always */
+	if (c->x86_vendor == X86_VENDOR_AMD && c->x86 == 15) {
+		/* disable GART TBL walk error reporting, which trips off 
+		   incorrectly with the IOMMU & 3ware & Cerberus. */
+		clear_bit(10, &bank[4]);
+		/* Lots of broken BIOS around that don't clear them
+		   by default and leave crap in there. Don't log. */
+		mce_bootlog = 0;
+	}
+
+}			
+
+static void __cpuinit mce_cpu_features(struct cpuinfo_x86 *c)
+{
+	switch (c->x86_vendor) {
+	case X86_VENDOR_INTEL:
+		mce_intel_feature_init(c);
+		break;
+	case X86_VENDOR_AMD:
+		mce_amd_feature_init(c);
+		break;
+	default:
+		break;
+	}
+}
+
+/* 
+ * Called for each booted CPU to set up machine checks.
+ * Must be called with preempt off. 
+ */
+void __cpuinit mcheck_init(struct cpuinfo_x86 *c)
+{
+	static cpumask_t mce_cpus = CPU_MASK_NONE;
+
+	mce_cpu_quirks(c); 
+
+	if (mce_dont_init ||
+	    cpu_test_and_set(smp_processor_id(), mce_cpus) ||
+	    !mce_available(c))
+		return;
+
+	mce_init(NULL);
+	mce_cpu_features(c);
+}
+
+/*
+ * Character device to read and clear the MCE log.
+ */
+
+static DEFINE_SPINLOCK(mce_state_lock);
+static int open_count;	/* #times opened */
+static int open_exclu;	/* already open exclusive? */
+
+static int mce_open(struct inode *inode, struct file *file)
+{
+	spin_lock(&mce_state_lock);
+
+	if (open_exclu || (open_count && (file->f_flags & O_EXCL))) {
+		spin_unlock(&mce_state_lock);
+		return -EBUSY;
+	}
+
+	if (file->f_flags & O_EXCL)
+		open_exclu = 1;
+	open_count++;
+
+	spin_unlock(&mce_state_lock);
+
+	return nonseekable_open(inode, file);
+}
+
+static int mce_release(struct inode *inode, struct file *file)
+{
+	spin_lock(&mce_state_lock);
+
+	open_count--;
+	open_exclu = 0;
+
+	spin_unlock(&mce_state_lock);
+
+	return 0;
+}
+
+static void collect_tscs(void *data) 
+{ 
+	unsigned long *cpu_tsc = (unsigned long *)data;
+	rdtscll(cpu_tsc[smp_processor_id()]);
+} 
+
+static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize, loff_t *off)
+{
+	unsigned long *cpu_tsc;
+	static DECLARE_MUTEX(mce_read_sem);
+	unsigned next;
+	char __user *buf = ubuf;
+	int i, err;
+
+	cpu_tsc = kmalloc(NR_CPUS * sizeof(long), GFP_KERNEL);
+	if (!cpu_tsc)
+		return -ENOMEM;
+
+	down(&mce_read_sem); 
+	next = rcu_dereference(mcelog.next);
+
+	/* Only supports full reads right now */
+	if (*off != 0 || usize < MCE_LOG_LEN*sizeof(struct mce)) { 
+		up(&mce_read_sem);
+		kfree(cpu_tsc);
+		return -EINVAL;
+	}
+
+	err = 0;
+	for (i = 0; i < next; i++) {		
+		unsigned long start = jiffies;
+		while (!mcelog.entry[i].finished) {
+			if (time_after_eq(jiffies, start + 2)) {
+				memset(mcelog.entry + i,0, sizeof(struct mce));
+				goto timeout;
+			}
+			cpu_relax();
+		}
+		smp_rmb();
+		err |= copy_to_user(buf, mcelog.entry + i, sizeof(struct mce));
+		buf += sizeof(struct mce); 
+ timeout:
+		;
+	} 
+
+	memset(mcelog.entry, 0, next * sizeof(struct mce));
+	mcelog.next = 0;
+
+	synchronize_sched();
+
+	/* Collect entries that were still getting written before the synchronize. */
+
+	on_each_cpu(collect_tscs, cpu_tsc, 1, 1);
+	for (i = next; i < MCE_LOG_LEN; i++) { 
+		if (mcelog.entry[i].finished && 
+		    mcelog.entry[i].tsc < cpu_tsc[mcelog.entry[i].cpu]) {  
+			err |= copy_to_user(buf, mcelog.entry+i, sizeof(struct mce));
+			smp_rmb();
+			buf += sizeof(struct mce);
+			memset(&mcelog.entry[i], 0, sizeof(struct mce));
+		}
+	} 	
+	up(&mce_read_sem);
+	kfree(cpu_tsc);
+	return err ? -EFAULT : buf - ubuf; 
+}
+
+static unsigned int mce_poll(struct file *file, poll_table *wait)
+{
+	poll_wait(file, &mce_wait, wait);
+	if (rcu_dereference(mcelog.next))
+		return POLLIN | POLLRDNORM;
+	return 0;
+}
+
+static int mce_ioctl(struct inode *i, struct file *f,unsigned int cmd, unsigned long arg)
+{
+	int __user *p = (int __user *)arg;
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM; 
+	switch (cmd) {
+	case MCE_GET_RECORD_LEN: 
+		return put_user(sizeof(struct mce), p);
+	case MCE_GET_LOG_LEN:
+		return put_user(MCE_LOG_LEN, p);		
+	case MCE_GETCLEAR_FLAGS: {
+		unsigned flags;
+		do { 
+			flags = mcelog.flags;
+		} while (cmpxchg(&mcelog.flags, flags, 0) != flags); 
+		return put_user(flags, p); 
+	}
+	default:
+		return -ENOTTY; 
+	} 
+}
+
+static const struct file_operations mce_chrdev_ops = {
+	.open = mce_open,
+	.release = mce_release,
+	.read = mce_read,
+	.poll = mce_poll,
+	.ioctl = mce_ioctl,
+};
+
+static struct miscdevice mce_log_device = {
+	MISC_MCELOG_MINOR,
+	"mcelog",
+	&mce_chrdev_ops,
+};
+
+static unsigned long old_cr4 __initdata;
+
+void __init stop_mce(void)
+{
+	old_cr4 = read_cr4();
+	clear_in_cr4(X86_CR4_MCE);
+}
+
+void __init restart_mce(void)
+{
+	if (old_cr4 & X86_CR4_MCE)
+		set_in_cr4(X86_CR4_MCE);
+}
+
+/* 
+ * Old style boot options parsing. Only for compatibility. 
+ */
+
+static int __init mcheck_disable(char *str)
+{
+	mce_dont_init = 1;
+	return 1;
+}
+
+/* mce=off disables machine check. Note you can reenable it later
+   using sysfs.
+   mce=TOLERANCELEVEL (number, see above)
+   mce=bootlog Log MCEs from before booting. Disabled by default on AMD.
+   mce=nobootlog Don't log MCEs from before booting. */
+static int __init mcheck_enable(char *str)
+{
+	if (*str == '=')
+		str++;
+	if (!strcmp(str, "off"))
+		mce_dont_init = 1;
+	else if (!strcmp(str, "bootlog") || !strcmp(str,"nobootlog"))
+		mce_bootlog = str[0] == 'b';
+	else if (isdigit(str[0]))
+		get_option(&str, &tolerant);
+	else
+		printk("mce= argument %s ignored. Please use /sys", str); 
+	return 1;
+}
+
+__setup("nomce", mcheck_disable);
+__setup("mce", mcheck_enable);
+
+/* 
+ * Sysfs support
+ */ 
+
+/* On resume clear all MCE state. Don't want to see leftovers from the BIOS.
+   Only one CPU is active at this time, the others get readded later using
+   CPU hotplug. */
+static int mce_resume(struct sys_device *dev)
+{
+	mce_init(NULL);
+	return 0;
+}
+
+/* Reinit MCEs after user configuration changes */
+static void mce_restart(void) 
+{ 
+	if (next_interval)
+		cancel_delayed_work(&mcheck_work);
+	/* Timer race is harmless here */
+	on_each_cpu(mce_init, NULL, 1, 1);       
+	next_interval = check_interval * HZ;
+	if (next_interval)
+		schedule_delayed_work(&mcheck_work,
+				      round_jiffies_relative(next_interval));
+}
+
+static struct sysdev_class mce_sysclass = {
+	.resume = mce_resume,
+	set_kset_name("machinecheck"),
+};
+
+DEFINE_PER_CPU(struct sys_device, device_mce);
+
+/* Why are there no generic functions for this? */
+#define ACCESSOR(name, var, start) \
+	static ssize_t show_ ## name(struct sys_device *s, char *buf) { 	   	   \
+		return sprintf(buf, "%lx\n", (unsigned long)var);		   \
+	} 									   \
+	static ssize_t set_ ## name(struct sys_device *s,const char *buf,size_t siz) { \
+		char *end; 							   \
+		unsigned long new = simple_strtoul(buf, &end, 0); 		   \
+		if (end == buf) return -EINVAL;					   \
+		var = new;							   \
+		start; 								   \
+		return end-buf;		     					   \
+	}									   \
+	static SYSDEV_ATTR(name, 0644, show_ ## name, set_ ## name);
+
+/* TBD should generate these dynamically based on number of available banks */
+ACCESSOR(bank0ctl,bank[0],mce_restart())
+ACCESSOR(bank1ctl,bank[1],mce_restart())
+ACCESSOR(bank2ctl,bank[2],mce_restart())
+ACCESSOR(bank3ctl,bank[3],mce_restart())
+ACCESSOR(bank4ctl,bank[4],mce_restart())
+ACCESSOR(bank5ctl,bank[5],mce_restart())
+
+static ssize_t show_trigger(struct sys_device *s, char *buf)
+{
+	strcpy(buf, trigger);
+	strcat(buf, "\n");
+	return strlen(trigger) + 1;
+}
+
+static ssize_t set_trigger(struct sys_device *s,const char *buf,size_t siz)
+{
+	char *p;
+	int len;
+	strncpy(trigger, buf, sizeof(trigger));
+	trigger[sizeof(trigger)-1] = 0;
+	len = strlen(trigger);
+	p = strchr(trigger, '\n');
+	if (*p) *p = 0;
+	return len;
+}
+
+static SYSDEV_ATTR(trigger, 0644, show_trigger, set_trigger);
+ACCESSOR(tolerant,tolerant,)
+ACCESSOR(check_interval,check_interval,mce_restart())
+static struct sysdev_attribute *mce_attributes[] = {
+	&attr_bank0ctl, &attr_bank1ctl, &attr_bank2ctl,
+	&attr_bank3ctl, &attr_bank4ctl, &attr_bank5ctl,
+	&attr_tolerant, &attr_check_interval, &attr_trigger,
+	NULL
+};
+
+/* Per cpu sysdev init.  All of the cpus still share the same ctl bank */
+static __cpuinit int mce_create_device(unsigned int cpu)
+{
+	int err;
+	int i;
+	if (!mce_available(&cpu_data[cpu]))
+		return -EIO;
+
+	per_cpu(device_mce,cpu).id = cpu;
+	per_cpu(device_mce,cpu).cls = &mce_sysclass;
+
+	err = sysdev_register(&per_cpu(device_mce,cpu));
+
+	if (!err) {
+		for (i = 0; mce_attributes[i]; i++)
+			sysdev_create_file(&per_cpu(device_mce,cpu),
+				mce_attributes[i]);
+	}
+	return err;
+}
+
+static void mce_remove_device(unsigned int cpu)
+{
+	int i;
+
+	for (i = 0; mce_attributes[i]; i++)
+		sysdev_remove_file(&per_cpu(device_mce,cpu),
+			mce_attributes[i]);
+	sysdev_unregister(&per_cpu(device_mce,cpu));
+	memset(&per_cpu(device_mce, cpu).kobj, 0, sizeof(struct kobject));
+}
+
+/* Get notified when a cpu comes on/off. Be hotplug friendly. */
+static int
+mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
+{
+	unsigned int cpu = (unsigned long)hcpu;
+
+	switch (action) {
+	case CPU_ONLINE:
+	case CPU_ONLINE_FROZEN:
+		mce_create_device(cpu);
+		break;
+	case CPU_DEAD:
+	case CPU_DEAD_FROZEN:
+		mce_remove_device(cpu);
+		break;
+	}
+	return NOTIFY_OK;
+}
+
+static struct notifier_block mce_cpu_notifier = {
+	.notifier_call = mce_cpu_callback,
+};
+
+static __init int mce_init_device(void)
+{
+	int err;
+	int i = 0;
+
+	if (!mce_available(&boot_cpu_data))
+		return -EIO;
+	err = sysdev_class_register(&mce_sysclass);
+
+	for_each_online_cpu(i) {
+		mce_create_device(i);
+	}
+
+	register_hotcpu_notifier(&mce_cpu_notifier);
+	misc_register(&mce_log_device);
+	return err;
+}
+
+device_initcall(mce_init_device);

arch/x86/kernel/mce_amd_64.c

@@ -0,0 +1,689 @@
+/*
+ *  (c) 2005, 2006 Advanced Micro Devices, Inc.
+ *  Your use of this code is subject to the terms and conditions of the
+ *  GNU general public license version 2. See "COPYING" or
+ *  http://www.gnu.org/licenses/gpl.html
+ *
+ *  Written by Jacob Shin - AMD, Inc.
+ *
+ *  Support : jacob.shin@amd.com
+ *
+ *  April 2006
+ *     - added support for AMD Family 0x10 processors
+ *
+ *  All MC4_MISCi registers are shared between multi-cores
+ */
+
+#include <linux/cpu.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/kobject.h>
+#include <linux/notifier.h>
+#include <linux/sched.h>
+#include <linux/smp.h>
+#include <linux/sysdev.h>
+#include <linux/sysfs.h>
+#include <asm/apic.h>
+#include <asm/mce.h>
+#include <asm/msr.h>
+#include <asm/percpu.h>
+#include <asm/idle.h>
+
+#define PFX               "mce_threshold: "
+#define VERSION           "version 1.1.1"
+#define NR_BANKS          6
+#define NR_BLOCKS         9
+#define THRESHOLD_MAX     0xFFF
+#define INT_TYPE_APIC     0x00020000
+#define MASK_VALID_HI     0x80000000
+#define MASK_CNTP_HI      0x40000000
+#define MASK_LOCKED_HI    0x20000000
+#define MASK_LVTOFF_HI    0x00F00000
+#define MASK_COUNT_EN_HI  0x00080000
+#define MASK_INT_TYPE_HI  0x00060000
+#define MASK_OVERFLOW_HI  0x00010000
+#define MASK_ERR_COUNT_HI 0x00000FFF
+#define MASK_BLKPTR_LO    0xFF000000
+#define MCG_XBLK_ADDR     0xC0000400
+
+struct threshold_block {
+	unsigned int block;
+	unsigned int bank;
+	unsigned int cpu;
+	u32 address;
+	u16 interrupt_enable;
+	u16 threshold_limit;
+	struct kobject kobj;
+	struct list_head miscj;
+};
+
+/* defaults used early on boot */
+static struct threshold_block threshold_defaults = {
+	.interrupt_enable = 0,
+	.threshold_limit = THRESHOLD_MAX,
+};
+
+struct threshold_bank {
+	struct kobject kobj;
+	struct threshold_block *blocks;
+	cpumask_t cpus;
+};
+static DEFINE_PER_CPU(struct threshold_bank *, threshold_banks[NR_BANKS]);
+
+#ifdef CONFIG_SMP
+static unsigned char shared_bank[NR_BANKS] = {
+	0, 0, 0, 0, 1
+};
+#endif
+
+static DEFINE_PER_CPU(unsigned char, bank_map);	/* see which banks are on */
+
+/*
+ * CPU Initialization
+ */
+
+/* must be called with correct cpu affinity */
+static void threshold_restart_bank(struct threshold_block *b,
+				   int reset, u16 old_limit)
+{
+	u32 mci_misc_hi, mci_misc_lo;
+
+	rdmsr(b->address, mci_misc_lo, mci_misc_hi);
+
+	if (b->threshold_limit < (mci_misc_hi & THRESHOLD_MAX))
+		reset = 1;	/* limit cannot be lower than err count */
+
+	if (reset) {		/* reset err count and overflow bit */
+		mci_misc_hi =
+		    (mci_misc_hi & ~(MASK_ERR_COUNT_HI | MASK_OVERFLOW_HI)) |
+		    (THRESHOLD_MAX - b->threshold_limit);
+	} else if (old_limit) {	/* change limit w/o reset */
+		int new_count = (mci_misc_hi & THRESHOLD_MAX) +
+		    (old_limit - b->threshold_limit);
+		mci_misc_hi = (mci_misc_hi & ~MASK_ERR_COUNT_HI) |
+		    (new_count & THRESHOLD_MAX);
+	}
+
+	b->interrupt_enable ?
+	    (mci_misc_hi = (mci_misc_hi & ~MASK_INT_TYPE_HI) | INT_TYPE_APIC) :
+	    (mci_misc_hi &= ~MASK_INT_TYPE_HI);
+
+	mci_misc_hi |= MASK_COUNT_EN_HI;
+	wrmsr(b->address, mci_misc_lo, mci_misc_hi);
+}
+
+/* cpu init entry point, called from mce.c with preempt off */
+void __cpuinit mce_amd_feature_init(struct cpuinfo_x86 *c)
+{
+	unsigned int bank, block;
+	unsigned int cpu = smp_processor_id();
+	u32 low = 0, high = 0, address = 0;
+
+	for (bank = 0; bank < NR_BANKS; ++bank) {
+		for (block = 0; block < NR_BLOCKS; ++block) {
+			if (block == 0)
+				address = MSR_IA32_MC0_MISC + bank * 4;
+			else if (block == 1) {
+				address = (low & MASK_BLKPTR_LO) >> 21;
+				if (!address)
+					break;
+				address += MCG_XBLK_ADDR;
+			}
+			else
+				++address;
+
+			if (rdmsr_safe(address, &low, &high))
+				break;
+
+			if (!(high & MASK_VALID_HI)) {
+				if (block)
+					continue;
+				else
+					break;
+			}
+
+			if (!(high & MASK_CNTP_HI)  ||
+			     (high & MASK_LOCKED_HI))
+				continue;
+
+			if (!block)
+				per_cpu(bank_map, cpu) |= (1 << bank);
+#ifdef CONFIG_SMP
+			if (shared_bank[bank] && c->cpu_core_id)
+				break;
+#endif
+			high &= ~MASK_LVTOFF_HI;
+			high |= K8_APIC_EXT_LVT_ENTRY_THRESHOLD << 20;
+			wrmsr(address, low, high);
+
+			setup_APIC_extended_lvt(K8_APIC_EXT_LVT_ENTRY_THRESHOLD,
+						THRESHOLD_APIC_VECTOR,
+						K8_APIC_EXT_INT_MSG_FIX, 0);
+
+			threshold_defaults.address = address;
+			threshold_restart_bank(&threshold_defaults, 0, 0);
+		}
+	}
+}
+
+/*
+ * APIC Interrupt Handler
+ */
+
+/*
+ * threshold interrupt handler will service THRESHOLD_APIC_VECTOR.
+ * the interrupt goes off when error_count reaches threshold_limit.
+ * the handler will simply log mcelog w/ software defined bank number.
+ */
+asmlinkage void mce_threshold_interrupt(void)
+{
+	unsigned int bank, block;
+	struct mce m;
+	u32 low = 0, high = 0, address = 0;
+
+	ack_APIC_irq();
+	exit_idle();
+	irq_enter();
+
+	memset(&m, 0, sizeof(m));
+	rdtscll(m.tsc);
+	m.cpu = smp_processor_id();
+
+	/* assume first bank caused it */
+	for (bank = 0; bank < NR_BANKS; ++bank) {
+		if (!(per_cpu(bank_map, m.cpu) & (1 << bank)))
+			continue;
+		for (block = 0; block < NR_BLOCKS; ++block) {
+			if (block == 0)
+				address = MSR_IA32_MC0_MISC + bank * 4;
+			else if (block == 1) {
+				address = (low & MASK_BLKPTR_LO) >> 21;
+				if (!address)
+					break;
+				address += MCG_XBLK_ADDR;
+			}
+			else
+				++address;
+
+			if (rdmsr_safe(address, &low, &high))
+				break;
+
+			if (!(high & MASK_VALID_HI)) {
+				if (block)
+					continue;
+				else
+					break;
+			}
+
+			if (!(high & MASK_CNTP_HI)  ||
+			     (high & MASK_LOCKED_HI))
+				continue;
+
+			/* Log the machine check that caused the threshold
+			   event. */
+			do_machine_check(NULL, 0);
+
+			if (high & MASK_OVERFLOW_HI) {
+				rdmsrl(address, m.misc);
+				rdmsrl(MSR_IA32_MC0_STATUS + bank * 4,
+				       m.status);
+				m.bank = K8_MCE_THRESHOLD_BASE
+				       + bank * NR_BLOCKS
+				       + block;
+				mce_log(&m);
+				goto out;
+			}
+		}
+	}
+out:
+	irq_exit();
+}
+
+/*
+ * Sysfs Interface
+ */
+
+struct threshold_attr {
+	struct attribute attr;
+	ssize_t(*show) (struct threshold_block *, char *);
+	ssize_t(*store) (struct threshold_block *, const char *, size_t count);
+};
+
+static cpumask_t affinity_set(unsigned int cpu)
+{
+	cpumask_t oldmask = current->cpus_allowed;
+	cpumask_t newmask = CPU_MASK_NONE;
+	cpu_set(cpu, newmask);
+	set_cpus_allowed(current, newmask);
+	return oldmask;
+}
+
+static void affinity_restore(cpumask_t oldmask)
+{
+	set_cpus_allowed(current, oldmask);
+}
+
+#define SHOW_FIELDS(name)                                           \
+static ssize_t show_ ## name(struct threshold_block * b, char *buf) \
+{                                                                   \
+        return sprintf(buf, "%lx\n", (unsigned long) b->name);      \
+}
+SHOW_FIELDS(interrupt_enable)
+SHOW_FIELDS(threshold_limit)
+
+static ssize_t store_interrupt_enable(struct threshold_block *b,
+				      const char *buf, size_t count)
+{
+	char *end;
+	cpumask_t oldmask;
+	unsigned long new = simple_strtoul(buf, &end, 0);
+	if (end == buf)
+		return -EINVAL;
+	b->interrupt_enable = !!new;
+
+	oldmask = affinity_set(b->cpu);
+	threshold_restart_bank(b, 0, 0);
+	affinity_restore(oldmask);
+
+	return end - buf;
+}
+
+static ssize_t store_threshold_limit(struct threshold_block *b,
+				     const char *buf, size_t count)
+{
+	char *end;
+	cpumask_t oldmask;
+	u16 old;
+	unsigned long new = simple_strtoul(buf, &end, 0);
+	if (end == buf)
+		return -EINVAL;
+	if (new > THRESHOLD_MAX)
+		new = THRESHOLD_MAX;
+	if (new < 1)
+		new = 1;
+	old = b->threshold_limit;
+	b->threshold_limit = new;
+
+	oldmask = affinity_set(b->cpu);
+	threshold_restart_bank(b, 0, old);
+	affinity_restore(oldmask);
+
+	return end - buf;
+}
+
+static ssize_t show_error_count(struct threshold_block *b, char *buf)
+{
+	u32 high, low;
+	cpumask_t oldmask;
+	oldmask = affinity_set(b->cpu);
+	rdmsr(b->address, low, high);
+	affinity_restore(oldmask);
+	return sprintf(buf, "%x\n",
+		       (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit));
+}
+
+static ssize_t store_error_count(struct threshold_block *b,
+				 const char *buf, size_t count)
+{
+	cpumask_t oldmask;
+	oldmask = affinity_set(b->cpu);
+	threshold_restart_bank(b, 1, 0);
+	affinity_restore(oldmask);
+	return 1;
+}
+
+#define THRESHOLD_ATTR(_name,_mode,_show,_store) {            \
+        .attr = {.name = __stringify(_name), .mode = _mode }, \
+        .show = _show,                                        \
+        .store = _store,                                      \
+};
+
+#define RW_ATTR(name)                                           \
+static struct threshold_attr name =                             \
+        THRESHOLD_ATTR(name, 0644, show_## name, store_## name)
+
+RW_ATTR(interrupt_enable);
+RW_ATTR(threshold_limit);
+RW_ATTR(error_count);
+
+static struct attribute *default_attrs[] = {
+	&interrupt_enable.attr,
+	&threshold_limit.attr,
+	&error_count.attr,
+	NULL
+};
+
+#define to_block(k) container_of(k, struct threshold_block, kobj)
+#define to_attr(a) container_of(a, struct threshold_attr, attr)
+
+static ssize_t show(struct kobject *kobj, struct attribute *attr, char *buf)
+{
+	struct threshold_block *b = to_block(kobj);
+	struct threshold_attr *a = to_attr(attr);
+	ssize_t ret;
+	ret = a->show ? a->show(b, buf) : -EIO;
+	return ret;
+}
+
+static ssize_t store(struct kobject *kobj, struct attribute *attr,
+		     const char *buf, size_t count)
+{
+	struct threshold_block *b = to_block(kobj);
+	struct threshold_attr *a = to_attr(attr);
+	ssize_t ret;
+	ret = a->store ? a->store(b, buf, count) : -EIO;
+	return ret;
+}
+
+static struct sysfs_ops threshold_ops = {
+	.show = show,
+	.store = store,
+};
+
+static struct kobj_type threshold_ktype = {
+	.sysfs_ops = &threshold_ops,
+	.default_attrs = default_attrs,
+};
+
+static __cpuinit int allocate_threshold_blocks(unsigned int cpu,
+					       unsigned int bank,
+					       unsigned int block,
+					       u32 address)
+{
+	int err;
+	u32 low, high;
+	struct threshold_block *b = NULL;
+
+	if ((bank >= NR_BANKS) || (block >= NR_BLOCKS))
+		return 0;
+
+	if (rdmsr_safe(address, &low, &high))
+		return 0;
+
+	if (!(high & MASK_VALID_HI)) {
+		if (block)
+			goto recurse;
+		else
+			return 0;
+	}
+
+	if (!(high & MASK_CNTP_HI)  ||
+	     (high & MASK_LOCKED_HI))
+		goto recurse;
+
+	b = kzalloc(sizeof(struct threshold_block), GFP_KERNEL);
+	if (!b)
+		return -ENOMEM;
+
+	b->block = block;
+	b->bank = bank;
+	b->cpu = cpu;
+	b->address = address;
+	b->interrupt_enable = 0;
+	b->threshold_limit = THRESHOLD_MAX;
+
+	INIT_LIST_HEAD(&b->miscj);
+
+	if (per_cpu(threshold_banks, cpu)[bank]->blocks)
+		list_add(&b->miscj,
+			 &per_cpu(threshold_banks, cpu)[bank]->blocks->miscj);
+	else
+		per_cpu(threshold_banks, cpu)[bank]->blocks = b;
+
+	kobject_set_name(&b->kobj, "misc%i", block);
+	b->kobj.parent = &per_cpu(threshold_banks, cpu)[bank]->kobj;
+	b->kobj.ktype = &threshold_ktype;
+	err = kobject_register(&b->kobj);
+	if (err)
+		goto out_free;
+recurse:
+	if (!block) {
+		address = (low & MASK_BLKPTR_LO) >> 21;
+		if (!address)
+			return 0;
+		address += MCG_XBLK_ADDR;
+	} else
+		++address;
+
+	err = allocate_threshold_blocks(cpu, bank, ++block, address);
+	if (err)
+		goto out_free;
+
+	return err;
+
+out_free:
+	if (b) {
+		kobject_unregister(&b->kobj);
+		kfree(b);
+	}
+	return err;
+}
+
+/* symlinks sibling shared banks to first core.  first core owns dir/files. */
+static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
+{
+	int i, err = 0;
+	struct threshold_bank *b = NULL;
+	cpumask_t oldmask = CPU_MASK_NONE;
+	char name[32];
+
+	sprintf(name, "threshold_bank%i", bank);
+
+#ifdef CONFIG_SMP
+	if (cpu_data[cpu].cpu_core_id && shared_bank[bank]) {	/* symlink */
+		i = first_cpu(cpu_core_map[cpu]);
+
+		/* first core not up yet */
+		if (cpu_data[i].cpu_core_id)
+			goto out;
+
+		/* already linked */
+		if (per_cpu(threshold_banks, cpu)[bank])
+			goto out;
+
+		b = per_cpu(threshold_banks, i)[bank];
+
+		if (!b)
+			goto out;
+
+		err = sysfs_create_link(&per_cpu(device_mce, cpu).kobj,
+					&b->kobj, name);
+		if (err)
+			goto out;
+
+		b->cpus = cpu_core_map[cpu];
+		per_cpu(threshold_banks, cpu)[bank] = b;
+		goto out;
+	}
+#endif
+
+	b = kzalloc(sizeof(struct threshold_bank), GFP_KERNEL);
+	if (!b) {
+		err = -ENOMEM;
+		goto out;
+	}
+
+	kobject_set_name(&b->kobj, "threshold_bank%i", bank);
+	b->kobj.parent = &per_cpu(device_mce, cpu).kobj;
+#ifndef CONFIG_SMP
+	b->cpus = CPU_MASK_ALL;
+#else
+	b->cpus = cpu_core_map[cpu];
+#endif
+	err = kobject_register(&b->kobj);
+	if (err)
+		goto out_free;
+
+	per_cpu(threshold_banks, cpu)[bank] = b;
+
+	oldmask = affinity_set(cpu);
+	err = allocate_threshold_blocks(cpu, bank, 0,
+					MSR_IA32_MC0_MISC + bank * 4);
+	affinity_restore(oldmask);
+
+	if (err)
+		goto out_free;
+
+	for_each_cpu_mask(i, b->cpus) {
+		if (i == cpu)
+			continue;
+
+		err = sysfs_create_link(&per_cpu(device_mce, i).kobj,
+					&b->kobj, name);
+		if (err)
+			goto out;
+
+		per_cpu(threshold_banks, i)[bank] = b;
+	}
+
+	goto out;
+
+out_free:
+	per_cpu(threshold_banks, cpu)[bank] = NULL;
+	kfree(b);
+out:
+	return err;
+}
+
+/* create dir/files for all valid threshold banks */
+static __cpuinit int threshold_create_device(unsigned int cpu)
+{
+	unsigned int bank;
+	int err = 0;
+
+	for (bank = 0; bank < NR_BANKS; ++bank) {
+		if (!(per_cpu(bank_map, cpu) & 1 << bank))
+			continue;
+		err = threshold_create_bank(cpu, bank);
+		if (err)
+			goto out;
+	}
+out:
+	return err;
+}
+
+/*
+ * let's be hotplug friendly.
+ * in case of multiple core processors, the first core always takes ownership
+ *   of shared sysfs dir/files, and rest of the cores will be symlinked to it.
+ */
+
+static void deallocate_threshold_block(unsigned int cpu,
+						 unsigned int bank)
+{
+	struct threshold_block *pos = NULL;
+	struct threshold_block *tmp = NULL;
+	struct threshold_bank *head = per_cpu(threshold_banks, cpu)[bank];
+
+	if (!head)
+		return;
+
+	list_for_each_entry_safe(pos, tmp, &head->blocks->miscj, miscj) {
+		kobject_unregister(&pos->kobj);
+		list_del(&pos->miscj);
+		kfree(pos);
+	}
+
+	kfree(per_cpu(threshold_banks, cpu)[bank]->blocks);
+	per_cpu(threshold_banks, cpu)[bank]->blocks = NULL;
+}
+
+static void threshold_remove_bank(unsigned int cpu, int bank)
+{
+	int i = 0;
+	struct threshold_bank *b;
+	char name[32];
+
+	b = per_cpu(threshold_banks, cpu)[bank];
+
+	if (!b)
+		return;
+
+	if (!b->blocks)
+		goto free_out;
+
+	sprintf(name, "threshold_bank%i", bank);
+
+#ifdef CONFIG_SMP
+	/* sibling symlink */
+	if (shared_bank[bank] && b->blocks->cpu != cpu) {
+		sysfs_remove_link(&per_cpu(device_mce, cpu).kobj, name);
+		per_cpu(threshold_banks, cpu)[bank] = NULL;
+		return;
+	}
+#endif
+
+	/* remove all sibling symlinks before unregistering */
+	for_each_cpu_mask(i, b->cpus) {
+		if (i == cpu)
+			continue;
+
+		sysfs_remove_link(&per_cpu(device_mce, i).kobj, name);
+		per_cpu(threshold_banks, i)[bank] = NULL;
+	}
+
+	deallocate_threshold_block(cpu, bank);
+
+free_out:
+	kobject_unregister(&b->kobj);
+	kfree(b);
+	per_cpu(threshold_banks, cpu)[bank] = NULL;
+}
+
+static void threshold_remove_device(unsigned int cpu)
+{
+	unsigned int bank;
+
+	for (bank = 0; bank < NR_BANKS; ++bank) {
+		if (!(per_cpu(bank_map, cpu) & 1 << bank))
+			continue;
+		threshold_remove_bank(cpu, bank);
+	}
+}
+
+/* get notified when a cpu comes on/off */
+static int threshold_cpu_callback(struct notifier_block *nfb,
+					    unsigned long action, void *hcpu)
+{
+	/* cpu was unsigned int to begin with */
+	unsigned int cpu = (unsigned long)hcpu;
+
+	if (cpu >= NR_CPUS)
+		goto out;
+
+	switch (action) {
+	case CPU_ONLINE:
+	case CPU_ONLINE_FROZEN:
+		threshold_create_device(cpu);
+		break;
+	case CPU_DEAD:
+	case CPU_DEAD_FROZEN:
+		threshold_remove_device(cpu);
+		break;
+	default:
+		break;
+	}
+      out:
+	return NOTIFY_OK;
+}
+
+static struct notifier_block threshold_cpu_notifier = {
+	.notifier_call = threshold_cpu_callback,
+};
+
+static __init int threshold_init_device(void)
+{
+	unsigned lcpu = 0;
+
+	/* to hit CPUs online before the notifier is up */
+	for_each_online_cpu(lcpu) {
+		int err = threshold_create_device(lcpu);
+		if (err)
+			return err;
+	}
+	register_hotcpu_notifier(&threshold_cpu_notifier);
+	return 0;
+}
+
+device_initcall(threshold_init_device);

arch/x86/kernel/mce_intel_64.c

@@ -0,0 +1,89 @@
+/*
+ * Intel specific MCE features.
+ * Copyright 2004 Zwane Mwaikambo <zwane@linuxpower.ca>
+ */
+
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/percpu.h>
+#include <asm/processor.h>
+#include <asm/msr.h>
+#include <asm/mce.h>
+#include <asm/hw_irq.h>
+#include <asm/idle.h>
+#include <asm/therm_throt.h>
+
+asmlinkage void smp_thermal_interrupt(void)
+{
+	__u64 msr_val;
+
+	ack_APIC_irq();
+
+	exit_idle();
+	irq_enter();
+
+	rdmsrl(MSR_IA32_THERM_STATUS, msr_val);
+	if (therm_throt_process(msr_val & 1))
+		mce_log_therm_throt_event(smp_processor_id(), msr_val);
+
+	irq_exit();
+}
+
+static void __cpuinit intel_init_thermal(struct cpuinfo_x86 *c)
+{
+	u32 l, h;
+	int tm2 = 0;
+	unsigned int cpu = smp_processor_id();
+
+	if (!cpu_has(c, X86_FEATURE_ACPI))
+		return;
+
+	if (!cpu_has(c, X86_FEATURE_ACC))
+		return;
+
+	/* first check if TM1 is already enabled by the BIOS, in which
+	 * case there might be some SMM goo which handles it, so we can't even
+	 * put a handler since it might be delivered via SMI already.
+	 */
+	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
+	h = apic_read(APIC_LVTTHMR);
+	if ((l & (1 << 3)) && (h & APIC_DM_SMI)) {
+		printk(KERN_DEBUG
+		       "CPU%d: Thermal monitoring handled by SMI\n", cpu);
+		return;
+	}
+
+	if (cpu_has(c, X86_FEATURE_TM2) && (l & (1 << 13)))
+		tm2 = 1;
+
+	if (h & APIC_VECTOR_MASK) {
+		printk(KERN_DEBUG
+		       "CPU%d: Thermal LVT vector (%#x) already "
+		       "installed\n", cpu, (h & APIC_VECTOR_MASK));
+		return;
+	}
+
+	h = THERMAL_APIC_VECTOR;
+	h |= (APIC_DM_FIXED | APIC_LVT_MASKED);
+	apic_write(APIC_LVTTHMR, h);
+
+	rdmsr(MSR_IA32_THERM_INTERRUPT, l, h);
+	wrmsr(MSR_IA32_THERM_INTERRUPT, l | 0x03, h);
+
+	rdmsr(MSR_IA32_MISC_ENABLE, l, h);
+	wrmsr(MSR_IA32_MISC_ENABLE, l | (1 << 3), h);
+
+	l = apic_read(APIC_LVTTHMR);
+	apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED);
+	printk(KERN_INFO "CPU%d: Thermal monitoring enabled (%s)\n",
+		cpu, tm2 ? "TM2" : "TM1");
+
+	/* enable thermal throttle processing */
+	atomic_set(&therm_throt_en, 1);
+	return;
+}
+
+void __cpuinit mce_intel_feature_init(struct cpuinfo_x86 *c)
+{
+	intel_init_thermal(c);
+}

arch/x86/kernel/module_64.c

@@ -0,0 +1,185 @@
+/*  Kernel module help for x86-64
+    Copyright (C) 2001 Rusty Russell.
+    Copyright (C) 2002,2003 Andi Kleen, SuSE Labs.
+
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+*/
+#include <linux/moduleloader.h>
+#include <linux/elf.h>
+#include <linux/vmalloc.h>
+#include <linux/fs.h>
+#include <linux/string.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/bug.h>
+
+#include <asm/system.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+
+#define DEBUGP(fmt...) 
+
+#ifndef CONFIG_UML
+void module_free(struct module *mod, void *module_region)
+{
+	vfree(module_region);
+	/* FIXME: If module_region == mod->init_region, trim exception
+           table entries. */
+}
+
+void *module_alloc(unsigned long size)
+{
+	struct vm_struct *area;
+
+	if (!size)
+		return NULL;
+	size = PAGE_ALIGN(size);
+	if (size > MODULES_LEN)
+		return NULL;
+
+	area = __get_vm_area(size, VM_ALLOC, MODULES_VADDR, MODULES_END);
+	if (!area)
+		return NULL;
+
+	return __vmalloc_area(area, GFP_KERNEL, PAGE_KERNEL_EXEC);
+}
+#endif
+
+/* We don't need anything special. */
+int module_frob_arch_sections(Elf_Ehdr *hdr,
+			      Elf_Shdr *sechdrs,
+			      char *secstrings,
+			      struct module *mod)
+{
+	return 0;
+}
+
+int apply_relocate_add(Elf64_Shdr *sechdrs,
+		   const char *strtab,
+		   unsigned int symindex,
+		   unsigned int relsec,
+		   struct module *me)
+{
+	unsigned int i;
+	Elf64_Rela *rel = (void *)sechdrs[relsec].sh_addr;
+	Elf64_Sym *sym;
+	void *loc;
+	u64 val; 
+
+	DEBUGP("Applying relocate section %u to %u\n", relsec,
+	       sechdrs[relsec].sh_info);
+	for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) {
+		/* This is where to make the change */
+		loc = (void *)sechdrs[sechdrs[relsec].sh_info].sh_addr
+			+ rel[i].r_offset;
+
+		/* This is the symbol it is referring to.  Note that all
+		   undefined symbols have been resolved.  */
+		sym = (Elf64_Sym *)sechdrs[symindex].sh_addr
+			+ ELF64_R_SYM(rel[i].r_info);
+
+	        DEBUGP("type %d st_value %Lx r_addend %Lx loc %Lx\n",
+		       (int)ELF64_R_TYPE(rel[i].r_info), 
+		       sym->st_value, rel[i].r_addend, (u64)loc);
+
+		val = sym->st_value + rel[i].r_addend; 
+
+		switch (ELF64_R_TYPE(rel[i].r_info)) {
+		case R_X86_64_NONE:
+			break;
+		case R_X86_64_64:
+			*(u64 *)loc = val;
+			break;
+		case R_X86_64_32:
+			*(u32 *)loc = val;
+			if (val != *(u32 *)loc)
+				goto overflow;
+			break;
+		case R_X86_64_32S:
+			*(s32 *)loc = val;
+			if ((s64)val != *(s32 *)loc)
+				goto overflow;
+			break;
+		case R_X86_64_PC32: 
+			val -= (u64)loc;
+			*(u32 *)loc = val;
+#if 0
+			if ((s64)val != *(s32 *)loc)
+				goto overflow; 
+#endif
+			break;
+		default:
+			printk(KERN_ERR "module %s: Unknown rela relocation: %Lu\n",
+			       me->name, ELF64_R_TYPE(rel[i].r_info));
+			return -ENOEXEC;
+		}
+	}
+	return 0;
+
+overflow:
+	printk(KERN_ERR "overflow in relocation type %d val %Lx\n", 
+	       (int)ELF64_R_TYPE(rel[i].r_info), val);
+	printk(KERN_ERR "`%s' likely not compiled with -mcmodel=kernel\n",
+	       me->name);
+	return -ENOEXEC;
+}
+
+int apply_relocate(Elf_Shdr *sechdrs,
+		   const char *strtab,
+		   unsigned int symindex,
+		   unsigned int relsec,
+		   struct module *me)
+{
+	printk("non add relocation not supported\n");
+	return -ENOSYS;
+} 
+
+int module_finalize(const Elf_Ehdr *hdr,
+                    const Elf_Shdr *sechdrs,
+                    struct module *me)
+{
+	const Elf_Shdr *s, *text = NULL, *alt = NULL, *locks = NULL;
+	char *secstrings = (void *)hdr + sechdrs[hdr->e_shstrndx].sh_offset;
+
+	for (s = sechdrs; s < sechdrs + hdr->e_shnum; s++) {
+		if (!strcmp(".text", secstrings + s->sh_name))
+			text = s;
+		if (!strcmp(".altinstructions", secstrings + s->sh_name))
+			alt = s;
+		if (!strcmp(".smp_locks", secstrings + s->sh_name))
+			locks= s;
+	}
+
+	if (alt) {
+		/* patch .altinstructions */
+		void *aseg = (void *)alt->sh_addr;
+		apply_alternatives(aseg, aseg + alt->sh_size);
+	}
+	if (locks && text) {
+		void *lseg = (void *)locks->sh_addr;
+		void *tseg = (void *)text->sh_addr;
+		alternatives_smp_module_add(me, me->name,
+					    lseg, lseg + locks->sh_size,
+					    tseg, tseg + text->sh_size);
+	}
+
+	return module_bug_finalize(hdr, sechdrs, me);
+}
+
+void module_arch_cleanup(struct module *mod)
+{
+	alternatives_smp_module_del(mod);
+	module_bug_cleanup(mod);
+}

arch/x86/kernel/mpparse_64.c

@@ -0,0 +1,852 @@
+/*
+ *	Intel Multiprocessor Specification 1.1 and 1.4
+ *	compliant MP-table parsing routines.
+ *
+ *	(c) 1995 Alan Cox, Building #3 <alan@redhat.com>
+ *	(c) 1998, 1999, 2000 Ingo Molnar <mingo@redhat.com>
+ *
+ *	Fixes
+ *		Erich Boleyn	:	MP v1.4 and additional changes.
+ *		Alan Cox	:	Added EBDA scanning
+ *		Ingo Molnar	:	various cleanups and rewrites
+ *		Maciej W. Rozycki:	Bits for default MP configurations
+ *		Paul Diefenbaugh:	Added full ACPI support
+ */
+
+#include <linux/mm.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/bootmem.h>
+#include <linux/kernel_stat.h>
+#include <linux/mc146818rtc.h>
+#include <linux/acpi.h>
+#include <linux/module.h>
+
+#include <asm/smp.h>
+#include <asm/mtrr.h>
+#include <asm/mpspec.h>
+#include <asm/pgalloc.h>
+#include <asm/io_apic.h>
+#include <asm/proto.h>
+#include <asm/acpi.h>
+
+/* Have we found an MP table */
+int smp_found_config;
+
+/*
+ * Various Linux-internal data structures created from the
+ * MP-table.
+ */
+DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES);
+int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
+
+static int mp_current_pci_id = 0;
+/* I/O APIC entries */
+struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];
+
+/* # of MP IRQ source entries */
+struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
+
+/* MP IRQ source entries */
+int mp_irq_entries;
+
+int nr_ioapics;
+unsigned long mp_lapic_addr = 0;
+
+
+
+/* Processor that is doing the boot up */
+unsigned int boot_cpu_id = -1U;
+/* Internal processor count */
+unsigned int num_processors __cpuinitdata = 0;
+
+unsigned disabled_cpus __cpuinitdata;
+
+/* Bitmask of physically existing CPUs */
+physid_mask_t phys_cpu_present_map = PHYSID_MASK_NONE;
+
+u8 bios_cpu_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID };
+
+
+/*
+ * Intel MP BIOS table parsing routines:
+ */
+
+/*
+ * Checksum an MP configuration block.
+ */
+
+static int __init mpf_checksum(unsigned char *mp, int len)
+{
+	int sum = 0;
+
+	while (len--)
+		sum += *mp++;
+
+	return sum & 0xFF;
+}
+
+static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
+{
+	int cpu;
+	cpumask_t tmp_map;
+	char *bootup_cpu = "";
+
+	if (!(m->mpc_cpuflag & CPU_ENABLED)) {
+		disabled_cpus++;
+		return;
+	}
+	if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
+		bootup_cpu = " (Bootup-CPU)";
+		boot_cpu_id = m->mpc_apicid;
+	}
+
+	printk(KERN_INFO "Processor #%d%s\n", m->mpc_apicid, bootup_cpu);
+
+	if (num_processors >= NR_CPUS) {
+		printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached."
+			" Processor ignored.\n", NR_CPUS);
+		return;
+	}
+
+	num_processors++;
+	cpus_complement(tmp_map, cpu_present_map);
+	cpu = first_cpu(tmp_map);
+
+	physid_set(m->mpc_apicid, phys_cpu_present_map);
+ 	if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
+ 		/*
+ 		 * bios_cpu_apicid is required to have processors listed
+ 		 * in same order as logical cpu numbers. Hence the first
+ 		 * entry is BSP, and so on.
+ 		 */
+		cpu = 0;
+ 	}
+	bios_cpu_apicid[cpu] = m->mpc_apicid;
+	x86_cpu_to_apicid[cpu] = m->mpc_apicid;
+
+	cpu_set(cpu, cpu_possible_map);
+	cpu_set(cpu, cpu_present_map);
+}
+
+static void __init MP_bus_info (struct mpc_config_bus *m)
+{
+	char str[7];
+
+	memcpy(str, m->mpc_bustype, 6);
+	str[6] = 0;
+	Dprintk("Bus #%d is %s\n", m->mpc_busid, str);
+
+	if (strncmp(str, "ISA", 3) == 0) {
+		set_bit(m->mpc_busid, mp_bus_not_pci);
+	} else if (strncmp(str, "PCI", 3) == 0) {
+		clear_bit(m->mpc_busid, mp_bus_not_pci);
+		mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id;
+		mp_current_pci_id++;
+	} else {
+		printk(KERN_ERR "Unknown bustype %s\n", str);
+	}
+}
+
+static int bad_ioapic(unsigned long address)
+{
+	if (nr_ioapics >= MAX_IO_APICS) {
+		printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded "
+			"(found %d)\n", MAX_IO_APICS, nr_ioapics);
+		panic("Recompile kernel with bigger MAX_IO_APICS!\n");
+	}
+	if (!address) {
+		printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address"
+			" found in table, skipping!\n");
+		return 1;
+	}
+	return 0;
+}
+
+static void __init MP_ioapic_info (struct mpc_config_ioapic *m)
+{
+	if (!(m->mpc_flags & MPC_APIC_USABLE))
+		return;
+
+	printk("I/O APIC #%d at 0x%X.\n",
+		m->mpc_apicid, m->mpc_apicaddr);
+
+	if (bad_ioapic(m->mpc_apicaddr))
+		return;
+
+	mp_ioapics[nr_ioapics] = *m;
+	nr_ioapics++;
+}
+
+static void __init MP_intsrc_info (struct mpc_config_intsrc *m)
+{
+	mp_irqs [mp_irq_entries] = *m;
+	Dprintk("Int: type %d, pol %d, trig %d, bus %d,"
+		" IRQ %02x, APIC ID %x, APIC INT %02x\n",
+			m->mpc_irqtype, m->mpc_irqflag & 3,
+			(m->mpc_irqflag >> 2) & 3, m->mpc_srcbus,
+			m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq);
+	if (++mp_irq_entries >= MAX_IRQ_SOURCES)
+		panic("Max # of irq sources exceeded!!\n");
+}
+
+static void __init MP_lintsrc_info (struct mpc_config_lintsrc *m)
+{
+	Dprintk("Lint: type %d, pol %d, trig %d, bus %d,"
+		" IRQ %02x, APIC ID %x, APIC LINT %02x\n",
+			m->mpc_irqtype, m->mpc_irqflag & 3,
+			(m->mpc_irqflag >> 2) &3, m->mpc_srcbusid,
+			m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint);
+}
+
+/*
+ * Read/parse the MPC
+ */
+
+static int __init smp_read_mpc(struct mp_config_table *mpc)
+{
+	char str[16];
+	int count=sizeof(*mpc);
+	unsigned char *mpt=((unsigned char *)mpc)+count;
+
+	if (memcmp(mpc->mpc_signature,MPC_SIGNATURE,4)) {
+		printk("MPTABLE: bad signature [%c%c%c%c]!\n",
+			mpc->mpc_signature[0],
+			mpc->mpc_signature[1],
+			mpc->mpc_signature[2],
+			mpc->mpc_signature[3]);
+		return 0;
+	}
+	if (mpf_checksum((unsigned char *)mpc,mpc->mpc_length)) {
+		printk("MPTABLE: checksum error!\n");
+		return 0;
+	}
+	if (mpc->mpc_spec!=0x01 && mpc->mpc_spec!=0x04) {
+		printk(KERN_ERR "MPTABLE: bad table version (%d)!!\n",
+			mpc->mpc_spec);
+		return 0;
+	}
+	if (!mpc->mpc_lapic) {
+		printk(KERN_ERR "MPTABLE: null local APIC address!\n");
+		return 0;
+	}
+	memcpy(str,mpc->mpc_oem,8);
+	str[8] = 0;
+	printk(KERN_INFO "MPTABLE: OEM ID: %s ",str);
+
+	memcpy(str,mpc->mpc_productid,12);
+	str[12] = 0;
+	printk("MPTABLE: Product ID: %s ",str);
+
+	printk("MPTABLE: APIC at: 0x%X\n",mpc->mpc_lapic);
+
+	/* save the local APIC address, it might be non-default */
+	if (!acpi_lapic)
+		mp_lapic_addr = mpc->mpc_lapic;
+
+	/*
+	 *	Now process the configuration blocks.
+	 */
+	while (count < mpc->mpc_length) {
+		switch(*mpt) {
+			case MP_PROCESSOR:
+			{
+				struct mpc_config_processor *m=
+					(struct mpc_config_processor *)mpt;
+				if (!acpi_lapic)
+					MP_processor_info(m);
+				mpt += sizeof(*m);
+				count += sizeof(*m);
+				break;
+			}
+			case MP_BUS:
+			{
+				struct mpc_config_bus *m=
+					(struct mpc_config_bus *)mpt;
+				MP_bus_info(m);
+				mpt += sizeof(*m);
+				count += sizeof(*m);
+				break;
+			}
+			case MP_IOAPIC:
+			{
+				struct mpc_config_ioapic *m=
+					(struct mpc_config_ioapic *)mpt;
+				MP_ioapic_info(m);
+				mpt += sizeof(*m);
+				count += sizeof(*m);
+				break;
+			}
+			case MP_INTSRC:
+			{
+				struct mpc_config_intsrc *m=
+					(struct mpc_config_intsrc *)mpt;
+
+				MP_intsrc_info(m);
+				mpt += sizeof(*m);
+				count += sizeof(*m);
+				break;
+			}
+			case MP_LINTSRC:
+			{
+				struct mpc_config_lintsrc *m=
+					(struct mpc_config_lintsrc *)mpt;
+				MP_lintsrc_info(m);
+				mpt += sizeof(*m);
+				count += sizeof(*m);
+				break;
+			}
+		}
+	}
+	setup_apic_routing();
+	if (!num_processors)
+		printk(KERN_ERR "MPTABLE: no processors registered!\n");
+	return num_processors;
+}
+
+static int __init ELCR_trigger(unsigned int irq)
+{
+	unsigned int port;
+
+	port = 0x4d0 + (irq >> 3);
+	return (inb(port) >> (irq & 7)) & 1;
+}
+
+static void __init construct_default_ioirq_mptable(int mpc_default_type)
+{
+	struct mpc_config_intsrc intsrc;
+	int i;
+	int ELCR_fallback = 0;
+
+	intsrc.mpc_type = MP_INTSRC;
+	intsrc.mpc_irqflag = 0;			/* conforming */
+	intsrc.mpc_srcbus = 0;
+	intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid;
+
+	intsrc.mpc_irqtype = mp_INT;
+
+	/*
+	 *  If true, we have an ISA/PCI system with no IRQ entries
+	 *  in the MP table. To prevent the PCI interrupts from being set up
+	 *  incorrectly, we try to use the ELCR. The sanity check to see if
+	 *  there is good ELCR data is very simple - IRQ0, 1, 2 and 13 can
+	 *  never be level sensitive, so we simply see if the ELCR agrees.
+	 *  If it does, we assume it's valid.
+	 */
+	if (mpc_default_type == 5) {
+		printk(KERN_INFO "ISA/PCI bus type with no IRQ information... falling back to ELCR\n");
+
+		if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) || ELCR_trigger(13))
+			printk(KERN_ERR "ELCR contains invalid data... not using ELCR\n");
+		else {
+			printk(KERN_INFO "Using ELCR to identify PCI interrupts\n");
+			ELCR_fallback = 1;
+		}
+	}
+
+	for (i = 0; i < 16; i++) {
+		switch (mpc_default_type) {
+		case 2:
+			if (i == 0 || i == 13)
+				continue;	/* IRQ0 & IRQ13 not connected */
+			/* fall through */
+		default:
+			if (i == 2)
+				continue;	/* IRQ2 is never connected */
+		}
+
+		if (ELCR_fallback) {
+			/*
+			 *  If the ELCR indicates a level-sensitive interrupt, we
+			 *  copy that information over to the MP table in the
+			 *  irqflag field (level sensitive, active high polarity).
+			 */
+			if (ELCR_trigger(i))
+				intsrc.mpc_irqflag = 13;
+			else
+				intsrc.mpc_irqflag = 0;
+		}
+
+		intsrc.mpc_srcbusirq = i;
+		intsrc.mpc_dstirq = i ? i : 2;		/* IRQ0 to INTIN2 */
+		MP_intsrc_info(&intsrc);
+	}
+
+	intsrc.mpc_irqtype = mp_ExtINT;
+	intsrc.mpc_srcbusirq = 0;
+	intsrc.mpc_dstirq = 0;				/* 8259A to INTIN0 */
+	MP_intsrc_info(&intsrc);
+}
+
+static inline void __init construct_default_ISA_mptable(int mpc_default_type)
+{
+	struct mpc_config_processor processor;
+	struct mpc_config_bus bus;
+	struct mpc_config_ioapic ioapic;
+	struct mpc_config_lintsrc lintsrc;
+	int linttypes[2] = { mp_ExtINT, mp_NMI };
+	int i;
+
+	/*
+	 * local APIC has default address
+	 */
+	mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
+
+	/*
+	 * 2 CPUs, numbered 0 & 1.
+	 */
+	processor.mpc_type = MP_PROCESSOR;
+	processor.mpc_apicver = 0;
+	processor.mpc_cpuflag = CPU_ENABLED;
+	processor.mpc_cpufeature = 0;
+	processor.mpc_featureflag = 0;
+	processor.mpc_reserved[0] = 0;
+	processor.mpc_reserved[1] = 0;
+	for (i = 0; i < 2; i++) {
+		processor.mpc_apicid = i;
+		MP_processor_info(&processor);
+	}
+
+	bus.mpc_type = MP_BUS;
+	bus.mpc_busid = 0;
+	switch (mpc_default_type) {
+		default:
+			printk(KERN_ERR "???\nUnknown standard configuration %d\n",
+				mpc_default_type);
+			/* fall through */
+		case 1:
+		case 5:
+			memcpy(bus.mpc_bustype, "ISA   ", 6);
+			break;
+	}
+	MP_bus_info(&bus);
+	if (mpc_default_type > 4) {
+		bus.mpc_busid = 1;
+		memcpy(bus.mpc_bustype, "PCI   ", 6);
+		MP_bus_info(&bus);
+	}
+
+	ioapic.mpc_type = MP_IOAPIC;
+	ioapic.mpc_apicid = 2;
+	ioapic.mpc_apicver = 0;
+	ioapic.mpc_flags = MPC_APIC_USABLE;
+	ioapic.mpc_apicaddr = 0xFEC00000;
+	MP_ioapic_info(&ioapic);
+
+	/*
+	 * We set up most of the low 16 IO-APIC pins according to MPS rules.
+	 */
+	construct_default_ioirq_mptable(mpc_default_type);
+
+	lintsrc.mpc_type = MP_LINTSRC;
+	lintsrc.mpc_irqflag = 0;		/* conforming */
+	lintsrc.mpc_srcbusid = 0;
+	lintsrc.mpc_srcbusirq = 0;
+	lintsrc.mpc_destapic = MP_APIC_ALL;
+	for (i = 0; i < 2; i++) {
+		lintsrc.mpc_irqtype = linttypes[i];
+		lintsrc.mpc_destapiclint = i;
+		MP_lintsrc_info(&lintsrc);
+	}
+}
+
+static struct intel_mp_floating *mpf_found;
+
+/*
+ * Scan the memory blocks for an SMP configuration block.
+ */
+void __init get_smp_config (void)
+{
+	struct intel_mp_floating *mpf = mpf_found;
+
+	/*
+ 	 * ACPI supports both logical (e.g. Hyper-Threading) and physical 
+ 	 * processors, where MPS only supports physical.
+ 	 */
+ 	if (acpi_lapic && acpi_ioapic) {
+ 		printk(KERN_INFO "Using ACPI (MADT) for SMP configuration information\n");
+ 		return;
+	}
+ 	else if (acpi_lapic)
+ 		printk(KERN_INFO "Using ACPI for processor (LAPIC) configuration information\n");
+
+	printk("Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification);
+
+	/*
+	 * Now see if we need to read further.
+	 */
+	if (mpf->mpf_feature1 != 0) {
+
+		printk(KERN_INFO "Default MP configuration #%d\n", mpf->mpf_feature1);
+		construct_default_ISA_mptable(mpf->mpf_feature1);
+
+	} else if (mpf->mpf_physptr) {
+
+		/*
+		 * Read the physical hardware table.  Anything here will
+		 * override the defaults.
+		 */
+		if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr))) {
+			smp_found_config = 0;
+			printk(KERN_ERR "BIOS bug, MP table errors detected!...\n");
+			printk(KERN_ERR "... disabling SMP support. (tell your hw vendor)\n");
+			return;
+		}
+		/*
+		 * If there are no explicit MP IRQ entries, then we are
+		 * broken.  We set up most of the low 16 IO-APIC pins to
+		 * ISA defaults and hope it will work.
+		 */
+		if (!mp_irq_entries) {
+			struct mpc_config_bus bus;
+
+			printk(KERN_ERR "BIOS bug, no explicit IRQ entries, using default mptable. (tell your hw vendor)\n");
+
+			bus.mpc_type = MP_BUS;
+			bus.mpc_busid = 0;
+			memcpy(bus.mpc_bustype, "ISA   ", 6);
+			MP_bus_info(&bus);
+
+			construct_default_ioirq_mptable(0);
+		}
+
+	} else
+		BUG();
+
+	printk(KERN_INFO "Processors: %d\n", num_processors);
+	/*
+	 * Only use the first configuration found.
+	 */
+}
+
+static int __init smp_scan_config (unsigned long base, unsigned long length)
+{
+	extern void __bad_mpf_size(void); 
+	unsigned int *bp = phys_to_virt(base);
+	struct intel_mp_floating *mpf;
+
+	Dprintk("Scan SMP from %p for %ld bytes.\n", bp,length);
+	if (sizeof(*mpf) != 16)
+		__bad_mpf_size();
+
+	while (length > 0) {
+		mpf = (struct intel_mp_floating *)bp;
+		if ((*bp == SMP_MAGIC_IDENT) &&
+			(mpf->mpf_length == 1) &&
+			!mpf_checksum((unsigned char *)bp, 16) &&
+			((mpf->mpf_specification == 1)
+				|| (mpf->mpf_specification == 4)) ) {
+
+			smp_found_config = 1;
+			reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE);
+			if (mpf->mpf_physptr)
+				reserve_bootmem_generic(mpf->mpf_physptr, PAGE_SIZE);
+			mpf_found = mpf;
+			return 1;
+		}
+		bp += 4;
+		length -= 16;
+	}
+	return 0;
+}
+
+void __init find_smp_config(void)
+{
+	unsigned int address;
+
+	/*
+	 * FIXME: Linux assumes you have 640K of base ram..
+	 * this continues the error...
+	 *
+	 * 1) Scan the bottom 1K for a signature
+	 * 2) Scan the top 1K of base RAM
+	 * 3) Scan the 64K of bios
+	 */
+	if (smp_scan_config(0x0,0x400) ||
+		smp_scan_config(639*0x400,0x400) ||
+			smp_scan_config(0xF0000,0x10000))
+		return;
+	/*
+	 * If it is an SMP machine we should know now.
+	 *
+	 * there is a real-mode segmented pointer pointing to the
+	 * 4K EBDA area at 0x40E, calculate and scan it here.
+	 *
+	 * NOTE! There are Linux loaders that will corrupt the EBDA
+	 * area, and as such this kind of SMP config may be less
+	 * trustworthy, simply because the SMP table may have been
+	 * stomped on during early boot. These loaders are buggy and
+	 * should be fixed.
+	 */
+
+	address = *(unsigned short *)phys_to_virt(0x40E);
+	address <<= 4;
+	if (smp_scan_config(address, 0x1000))
+		return;
+
+	/* If we have come this far, we did not find an MP table  */
+	 printk(KERN_INFO "No mptable found.\n");
+}
+
+/* --------------------------------------------------------------------------
+                            ACPI-based MP Configuration
+   -------------------------------------------------------------------------- */
+
+#ifdef CONFIG_ACPI
+
+void __init mp_register_lapic_address(u64 address)
+{
+	mp_lapic_addr = (unsigned long) address;
+	set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr);
+	if (boot_cpu_id == -1U)
+		boot_cpu_id = GET_APIC_ID(apic_read(APIC_ID));
+}
+
+void __cpuinit mp_register_lapic (u8 id, u8 enabled)
+{
+	struct mpc_config_processor processor;
+	int			boot_cpu = 0;
+	
+	if (id == boot_cpu_id)
+		boot_cpu = 1;
+
+	processor.mpc_type = MP_PROCESSOR;
+	processor.mpc_apicid = id;
+	processor.mpc_apicver = 0;
+	processor.mpc_cpuflag = (enabled ? CPU_ENABLED : 0);
+	processor.mpc_cpuflag |= (boot_cpu ? CPU_BOOTPROCESSOR : 0);
+	processor.mpc_cpufeature = 0;
+	processor.mpc_featureflag = 0;
+	processor.mpc_reserved[0] = 0;
+	processor.mpc_reserved[1] = 0;
+
+	MP_processor_info(&processor);
+}
+
+#define MP_ISA_BUS		0
+#define MP_MAX_IOAPIC_PIN	127
+
+static struct mp_ioapic_routing {
+	int			apic_id;
+	int			gsi_start;
+	int			gsi_end;
+	u32			pin_programmed[4];
+} mp_ioapic_routing[MAX_IO_APICS];
+
+static int mp_find_ioapic(int gsi)
+{
+	int i = 0;
+
+	/* Find the IOAPIC that manages this GSI. */
+	for (i = 0; i < nr_ioapics; i++) {
+		if ((gsi >= mp_ioapic_routing[i].gsi_start)
+			&& (gsi <= mp_ioapic_routing[i].gsi_end))
+			return i;
+	}
+
+	printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
+	return -1;
+}
+
+static u8 uniq_ioapic_id(u8 id)
+{
+	int i;
+	DECLARE_BITMAP(used, 256);
+	bitmap_zero(used, 256);
+	for (i = 0; i < nr_ioapics; i++) {
+		struct mpc_config_ioapic *ia = &mp_ioapics[i];
+		__set_bit(ia->mpc_apicid, used);
+	}
+	if (!test_bit(id, used))
+		return id;
+	return find_first_zero_bit(used, 256);
+}
+
+void __init mp_register_ioapic(u8 id, u32 address, u32 gsi_base)
+{
+	int idx = 0;
+
+	if (bad_ioapic(address))
+		return;
+
+	idx = nr_ioapics;
+
+	mp_ioapics[idx].mpc_type = MP_IOAPIC;
+	mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE;
+	mp_ioapics[idx].mpc_apicaddr = address;
+
+	set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
+	mp_ioapics[idx].mpc_apicid = uniq_ioapic_id(id);
+	mp_ioapics[idx].mpc_apicver = 0;
+	
+	/* 
+	 * Build basic IRQ lookup table to facilitate gsi->io_apic lookups
+	 * and to prevent reprogramming of IOAPIC pins (PCI IRQs).
+	 */
+	mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid;
+	mp_ioapic_routing[idx].gsi_start = gsi_base;
+	mp_ioapic_routing[idx].gsi_end = gsi_base + 
+		io_apic_get_redir_entries(idx);
+
+	printk(KERN_INFO "IOAPIC[%d]: apic_id %d, address 0x%x, "
+		"GSI %d-%d\n", idx, mp_ioapics[idx].mpc_apicid, 
+		mp_ioapics[idx].mpc_apicaddr,
+		mp_ioapic_routing[idx].gsi_start,
+		mp_ioapic_routing[idx].gsi_end);
+
+	nr_ioapics++;
+}
+
+void __init
+mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32	gsi)
+{
+	struct mpc_config_intsrc intsrc;
+	int			ioapic = -1;
+	int			pin = -1;
+
+	/* 
+	 * Convert 'gsi' to 'ioapic.pin'.
+	 */
+	ioapic = mp_find_ioapic(gsi);
+	if (ioapic < 0)
+		return;
+	pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
+
+	/*
+	 * TBD: This check is for faulty timer entries, where the override
+	 *      erroneously sets the trigger to level, resulting in a HUGE 
+	 *      increase of timer interrupts!
+	 */
+	if ((bus_irq == 0) && (trigger == 3))
+		trigger = 1;
+
+	intsrc.mpc_type = MP_INTSRC;
+	intsrc.mpc_irqtype = mp_INT;
+	intsrc.mpc_irqflag = (trigger << 2) | polarity;
+	intsrc.mpc_srcbus = MP_ISA_BUS;
+	intsrc.mpc_srcbusirq = bus_irq;				       /* IRQ */
+	intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid;	   /* APIC ID */
+	intsrc.mpc_dstirq = pin;				    /* INTIN# */
+
+	Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, %d-%d\n", 
+		intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, 
+		(intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, 
+		intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq);
+
+	mp_irqs[mp_irq_entries] = intsrc;
+	if (++mp_irq_entries == MAX_IRQ_SOURCES)
+		panic("Max # of irq sources exceeded!\n");
+}
+
+void __init mp_config_acpi_legacy_irqs(void)
+{
+	struct mpc_config_intsrc intsrc;
+	int i = 0;
+	int ioapic = -1;
+
+	/* 
+	 * Fabricate the legacy ISA bus (bus #31).
+	 */
+	set_bit(MP_ISA_BUS, mp_bus_not_pci);
+
+	/* 
+	 * Locate the IOAPIC that manages the ISA IRQs (0-15). 
+	 */
+	ioapic = mp_find_ioapic(0);
+	if (ioapic < 0)
+		return;
+
+	intsrc.mpc_type = MP_INTSRC;
+	intsrc.mpc_irqflag = 0;					/* Conforming */
+	intsrc.mpc_srcbus = MP_ISA_BUS;
+	intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid;
+
+	/* 
+	 * Use the default configuration for the IRQs 0-15.  Unless
+	 * overridden by (MADT) interrupt source override entries.
+	 */
+	for (i = 0; i < 16; i++) {
+		int idx;
+
+		for (idx = 0; idx < mp_irq_entries; idx++) {
+			struct mpc_config_intsrc *irq = mp_irqs + idx;
+
+			/* Do we already have a mapping for this ISA IRQ? */
+			if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i)
+				break;
+
+			/* Do we already have a mapping for this IOAPIC pin */
+			if ((irq->mpc_dstapic == intsrc.mpc_dstapic) &&
+				(irq->mpc_dstirq == i))
+				break;
+		}
+
+		if (idx != mp_irq_entries) {
+			printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i);
+			continue;			/* IRQ already used */
+		}
+
+		intsrc.mpc_irqtype = mp_INT;
+		intsrc.mpc_srcbusirq = i;		   /* Identity mapped */
+		intsrc.mpc_dstirq = i;
+
+		Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, "
+			"%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, 
+			(intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, 
+			intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, 
+			intsrc.mpc_dstirq);
+
+		mp_irqs[mp_irq_entries] = intsrc;
+		if (++mp_irq_entries == MAX_IRQ_SOURCES)
+			panic("Max # of irq sources exceeded!\n");
+	}
+}
+
+int mp_register_gsi(u32 gsi, int triggering, int polarity)
+{
+	int ioapic = -1;
+	int ioapic_pin = 0;
+	int idx, bit = 0;
+
+	if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC)
+		return gsi;
+
+	/* Don't set up the ACPI SCI because it's already set up */
+	if (acpi_gbl_FADT.sci_interrupt == gsi)
+		return gsi;
+
+	ioapic = mp_find_ioapic(gsi);
+	if (ioapic < 0) {
+		printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi);
+		return gsi;
+	}
+
+	ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_start;
+
+	/* 
+	 * Avoid pin reprogramming.  PRTs typically include entries  
+	 * with redundant pin->gsi mappings (but unique PCI devices);
+	 * we only program the IOAPIC on the first.
+	 */
+	bit = ioapic_pin % 32;
+	idx = (ioapic_pin < 32) ? 0 : (ioapic_pin / 32);
+	if (idx > 3) {
+		printk(KERN_ERR "Invalid reference to IOAPIC pin "
+			"%d-%d\n", mp_ioapic_routing[ioapic].apic_id, 
+			ioapic_pin);
+		return gsi;
+	}
+	if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) {
+		Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n",
+			mp_ioapic_routing[ioapic].apic_id, ioapic_pin);
+		return gsi;
+	}
+
+	mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit);
+
+	io_apic_set_pci_routing(ioapic, ioapic_pin, gsi,
+		triggering == ACPI_EDGE_SENSITIVE ? 0 : 1,
+		polarity == ACPI_ACTIVE_HIGH ? 0 : 1);
+	return gsi;
+}
+#endif /*CONFIG_ACPI*/

arch/x86/kernel/nmi_64.c

@@ -0,0 +1,483 @@
+/*
+ *  linux/arch/x86_64/nmi.c
+ *
+ *  NMI watchdog support on APIC systems
+ *
+ *  Started by Ingo Molnar <mingo@redhat.com>
+ *
+ *  Fixes:
+ *  Mikael Pettersson	: AMD K7 support for local APIC NMI watchdog.
+ *  Mikael Pettersson	: Power Management for local APIC NMI watchdog.
+ *  Pavel Machek and
+ *  Mikael Pettersson	: PM converted to driver model. Disable/enable API.
+ */
+
+#include <linux/nmi.h>
+#include <linux/mm.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/sysdev.h>
+#include <linux/sysctl.h>
+#include <linux/kprobes.h>
+#include <linux/cpumask.h>
+#include <linux/kdebug.h>
+
+#include <asm/smp.h>
+#include <asm/nmi.h>
+#include <asm/proto.h>
+#include <asm/mce.h>
+
+int unknown_nmi_panic;
+int nmi_watchdog_enabled;
+int panic_on_unrecovered_nmi;
+
+static cpumask_t backtrace_mask = CPU_MASK_NONE;
+
+/* nmi_active:
+ * >0: the lapic NMI watchdog is active, but can be disabled
+ * <0: the lapic NMI watchdog has not been set up, and cannot
+ *     be enabled
+ *  0: the lapic NMI watchdog is disabled, but can be enabled
+ */
+atomic_t nmi_active = ATOMIC_INIT(0);		/* oprofile uses this */
+int panic_on_timeout;
+
+unsigned int nmi_watchdog = NMI_DEFAULT;
+static unsigned int nmi_hz = HZ;
+
+static DEFINE_PER_CPU(short, wd_enabled);
+
+/* local prototypes */
+static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu);
+
+/* Run after command line and cpu_init init, but before all other checks */
+void nmi_watchdog_default(void)
+{
+	if (nmi_watchdog != NMI_DEFAULT)
+		return;
+	nmi_watchdog = NMI_NONE;
+}
+
+static int endflag __initdata = 0;
+
+#ifdef CONFIG_SMP
+/* The performance counters used by NMI_LOCAL_APIC don't trigger when
+ * the CPU is idle. To make sure the NMI watchdog really ticks on all
+ * CPUs during the test make them busy.
+ */
+static __init void nmi_cpu_busy(void *data)
+{
+	local_irq_enable_in_hardirq();
+	/* Intentionally don't use cpu_relax here. This is
+	   to make sure that the performance counter really ticks,
+	   even if there is a simulator or similar that catches the
+	   pause instruction. On a real HT machine this is fine because
+	   all other CPUs are busy with "useless" delay loops and don't
+	   care if they get somewhat less cycles. */
+	while (endflag == 0)
+		mb();
+}
+#endif
+
+int __init check_nmi_watchdog (void)
+{
+	int *counts;
+	int cpu;
+
+	if ((nmi_watchdog == NMI_NONE) || (nmi_watchdog == NMI_DISABLED)) 
+		return 0;
+
+	if (!atomic_read(&nmi_active))
+		return 0;
+
+	counts = kmalloc(NR_CPUS * sizeof(int), GFP_KERNEL);
+	if (!counts)
+		return -1;
+
+	printk(KERN_INFO "testing NMI watchdog ... ");
+
+#ifdef CONFIG_SMP
+	if (nmi_watchdog == NMI_LOCAL_APIC)
+		smp_call_function(nmi_cpu_busy, (void *)&endflag, 0, 0);
+#endif
+
+	for (cpu = 0; cpu < NR_CPUS; cpu++)
+		counts[cpu] = cpu_pda(cpu)->__nmi_count;
+	local_irq_enable();
+	mdelay((20*1000)/nmi_hz); // wait 20 ticks
+
+	for_each_online_cpu(cpu) {
+		if (!per_cpu(wd_enabled, cpu))
+			continue;
+		if (cpu_pda(cpu)->__nmi_count - counts[cpu] <= 5) {
+			printk("CPU#%d: NMI appears to be stuck (%d->%d)!\n",
+			       cpu,
+			       counts[cpu],
+			       cpu_pda(cpu)->__nmi_count);
+			per_cpu(wd_enabled, cpu) = 0;
+			atomic_dec(&nmi_active);
+		}
+	}
+	if (!atomic_read(&nmi_active)) {
+		kfree(counts);
+		atomic_set(&nmi_active, -1);
+		endflag = 1;
+		return -1;
+	}
+	endflag = 1;
+	printk("OK.\n");
+
+	/* now that we know it works we can reduce NMI frequency to
+	   something more reasonable; makes a difference in some configs */
+	if (nmi_watchdog == NMI_LOCAL_APIC)
+		nmi_hz = lapic_adjust_nmi_hz(1);
+
+	kfree(counts);
+	return 0;
+}
+
+int __init setup_nmi_watchdog(char *str)
+{
+	int nmi;
+
+	if (!strncmp(str,"panic",5)) {
+		panic_on_timeout = 1;
+		str = strchr(str, ',');
+		if (!str)
+			return 1;
+		++str;
+	}
+
+	get_option(&str, &nmi);
+
+	if ((nmi >= NMI_INVALID) || (nmi < NMI_NONE))
+		return 0;
+
+	nmi_watchdog = nmi;
+	return 1;
+}
+
+__setup("nmi_watchdog=", setup_nmi_watchdog);
+
+
+static void __acpi_nmi_disable(void *__unused)
+{
+	apic_write(APIC_LVT0, APIC_DM_NMI | APIC_LVT_MASKED);
+}
+
+/*
+ * Disable timer based NMIs on all CPUs:
+ */
+void acpi_nmi_disable(void)
+{
+	if (atomic_read(&nmi_active) && nmi_watchdog == NMI_IO_APIC)
+		on_each_cpu(__acpi_nmi_disable, NULL, 0, 1);
+}
+
+static void __acpi_nmi_enable(void *__unused)
+{
+	apic_write(APIC_LVT0, APIC_DM_NMI);
+}
+
+/*
+ * Enable timer based NMIs on all CPUs:
+ */
+void acpi_nmi_enable(void)
+{
+	if (atomic_read(&nmi_active) && nmi_watchdog == NMI_IO_APIC)
+		on_each_cpu(__acpi_nmi_enable, NULL, 0, 1);
+}
+#ifdef CONFIG_PM
+
+static int nmi_pm_active; /* nmi_active before suspend */
+
+static int lapic_nmi_suspend(struct sys_device *dev, pm_message_t state)
+{
+	/* only CPU0 goes here, other CPUs should be offline */
+	nmi_pm_active = atomic_read(&nmi_active);
+	stop_apic_nmi_watchdog(NULL);
+	BUG_ON(atomic_read(&nmi_active) != 0);
+	return 0;
+}
+
+static int lapic_nmi_resume(struct sys_device *dev)
+{
+	/* only CPU0 goes here, other CPUs should be offline */
+	if (nmi_pm_active > 0) {
+		setup_apic_nmi_watchdog(NULL);
+		touch_nmi_watchdog();
+	}
+	return 0;
+}
+
+static struct sysdev_class nmi_sysclass = {
+	set_kset_name("lapic_nmi"),
+	.resume		= lapic_nmi_resume,
+	.suspend	= lapic_nmi_suspend,
+};
+
+static struct sys_device device_lapic_nmi = {
+	.id		= 0,
+	.cls	= &nmi_sysclass,
+};
+
+static int __init init_lapic_nmi_sysfs(void)
+{
+	int error;
+
+	/* should really be a BUG_ON but b/c this is an
+	 * init call, it just doesn't work.  -dcz
+	 */
+	if (nmi_watchdog != NMI_LOCAL_APIC)
+		return 0;
+
+	if ( atomic_read(&nmi_active) < 0 )
+		return 0;
+
+	error = sysdev_class_register(&nmi_sysclass);
+	if (!error)
+		error = sysdev_register(&device_lapic_nmi);
+	return error;
+}
+/* must come after the local APIC's device_initcall() */
+late_initcall(init_lapic_nmi_sysfs);
+
+#endif	/* CONFIG_PM */
+
+void setup_apic_nmi_watchdog(void *unused)
+{
+	if (__get_cpu_var(wd_enabled) == 1)
+		return;
+
+	/* cheap hack to support suspend/resume */
+	/* if cpu0 is not active neither should the other cpus */
+	if ((smp_processor_id() != 0) && (atomic_read(&nmi_active) <= 0))
+		return;
+
+	switch (nmi_watchdog) {
+	case NMI_LOCAL_APIC:
+		__get_cpu_var(wd_enabled) = 1;
+		if (lapic_watchdog_init(nmi_hz) < 0) {
+			__get_cpu_var(wd_enabled) = 0;
+			return;
+		}
+		/* FALL THROUGH */
+	case NMI_IO_APIC:
+		__get_cpu_var(wd_enabled) = 1;
+		atomic_inc(&nmi_active);
+	}
+}
+
+void stop_apic_nmi_watchdog(void *unused)
+{
+	/* only support LOCAL and IO APICs for now */
+	if ((nmi_watchdog != NMI_LOCAL_APIC) &&
+	    (nmi_watchdog != NMI_IO_APIC))
+	    	return;
+	if (__get_cpu_var(wd_enabled) == 0)
+		return;
+	if (nmi_watchdog == NMI_LOCAL_APIC)
+		lapic_watchdog_stop();
+	__get_cpu_var(wd_enabled) = 0;
+	atomic_dec(&nmi_active);
+}
+
+/*
+ * the best way to detect whether a CPU has a 'hard lockup' problem
+ * is to check it's local APIC timer IRQ counts. If they are not
+ * changing then that CPU has some problem.
+ *
+ * as these watchdog NMI IRQs are generated on every CPU, we only
+ * have to check the current processor.
+ */
+
+static DEFINE_PER_CPU(unsigned, last_irq_sum);
+static DEFINE_PER_CPU(local_t, alert_counter);
+static DEFINE_PER_CPU(int, nmi_touch);
+
+void touch_nmi_watchdog(void)
+{
+	if (nmi_watchdog > 0) {
+		unsigned cpu;
+
+		/*
+ 		 * Tell other CPUs to reset their alert counters. We cannot
+		 * do it ourselves because the alert count increase is not
+		 * atomic.
+		 */
+		for_each_present_cpu(cpu) {
+			if (per_cpu(nmi_touch, cpu) != 1)
+				per_cpu(nmi_touch, cpu) = 1;
+		}
+	}
+
+ 	touch_softlockup_watchdog();
+}
+
+int __kprobes nmi_watchdog_tick(struct pt_regs * regs, unsigned reason)
+{
+	int sum;
+	int touched = 0;
+	int cpu = smp_processor_id();
+	int rc = 0;
+
+	/* check for other users first */
+	if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT)
+			== NOTIFY_STOP) {
+		rc = 1;
+		touched = 1;
+	}
+
+	sum = read_pda(apic_timer_irqs);
+	if (__get_cpu_var(nmi_touch)) {
+		__get_cpu_var(nmi_touch) = 0;
+		touched = 1;
+	}
+
+	if (cpu_isset(cpu, backtrace_mask)) {
+		static DEFINE_SPINLOCK(lock);	/* Serialise the printks */
+
+		spin_lock(&lock);
+		printk("NMI backtrace for cpu %d\n", cpu);
+		dump_stack();
+		spin_unlock(&lock);
+		cpu_clear(cpu, backtrace_mask);
+	}
+
+#ifdef CONFIG_X86_MCE
+	/* Could check oops_in_progress here too, but it's safer
+	   not too */
+	if (atomic_read(&mce_entry) > 0)
+		touched = 1;
+#endif
+	/* if the apic timer isn't firing, this cpu isn't doing much */
+	if (!touched && __get_cpu_var(last_irq_sum) == sum) {
+		/*
+		 * Ayiee, looks like this CPU is stuck ...
+		 * wait a few IRQs (5 seconds) before doing the oops ...
+		 */
+		local_inc(&__get_cpu_var(alert_counter));
+		if (local_read(&__get_cpu_var(alert_counter)) == 5*nmi_hz)
+			die_nmi("NMI Watchdog detected LOCKUP on CPU %d\n", regs,
+				panic_on_timeout);
+	} else {
+		__get_cpu_var(last_irq_sum) = sum;
+		local_set(&__get_cpu_var(alert_counter), 0);
+	}
+
+	/* see if the nmi watchdog went off */
+	if (!__get_cpu_var(wd_enabled))
+		return rc;
+	switch (nmi_watchdog) {
+	case NMI_LOCAL_APIC:
+		rc |= lapic_wd_event(nmi_hz);
+		break;
+	case NMI_IO_APIC:
+		/* don't know how to accurately check for this.
+		 * just assume it was a watchdog timer interrupt
+		 * This matches the old behaviour.
+		 */
+		rc = 1;
+		break;
+	}
+	return rc;
+}
+
+static unsigned ignore_nmis;
+
+asmlinkage __kprobes void do_nmi(struct pt_regs * regs, long error_code)
+{
+	nmi_enter();
+	add_pda(__nmi_count,1);
+	if (!ignore_nmis)
+		default_do_nmi(regs);
+	nmi_exit();
+}
+
+int do_nmi_callback(struct pt_regs * regs, int cpu)
+{
+#ifdef CONFIG_SYSCTL
+	if (unknown_nmi_panic)
+		return unknown_nmi_panic_callback(regs, cpu);
+#endif
+	return 0;
+}
+
+void stop_nmi(void)
+{
+	acpi_nmi_disable();
+	ignore_nmis++;
+}
+
+void restart_nmi(void)
+{
+	ignore_nmis--;
+	acpi_nmi_enable();
+}
+
+#ifdef CONFIG_SYSCTL
+
+static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu)
+{
+	unsigned char reason = get_nmi_reason();
+	char buf[64];
+
+	sprintf(buf, "NMI received for unknown reason %02x\n", reason);
+	die_nmi(buf, regs, 1);	/* Always panic here */
+	return 0;
+}
+
+/*
+ * proc handler for /proc/sys/kernel/nmi
+ */
+int proc_nmi_enabled(struct ctl_table *table, int write, struct file *file,
+			void __user *buffer, size_t *length, loff_t *ppos)
+{
+	int old_state;
+
+	nmi_watchdog_enabled = (atomic_read(&nmi_active) > 0) ? 1 : 0;
+	old_state = nmi_watchdog_enabled;
+	proc_dointvec(table, write, file, buffer, length, ppos);
+	if (!!old_state == !!nmi_watchdog_enabled)
+		return 0;
+
+	if (atomic_read(&nmi_active) < 0 || nmi_watchdog == NMI_DISABLED) {
+		printk( KERN_WARNING "NMI watchdog is permanently disabled\n");
+		return -EIO;
+	}
+
+	/* if nmi_watchdog is not set yet, then set it */
+	nmi_watchdog_default();
+
+	if (nmi_watchdog == NMI_LOCAL_APIC) {
+		if (nmi_watchdog_enabled)
+			enable_lapic_nmi_watchdog();
+		else
+			disable_lapic_nmi_watchdog();
+	} else {
+		printk( KERN_WARNING
+			"NMI watchdog doesn't know what hardware to touch\n");
+		return -EIO;
+	}
+	return 0;
+}
+
+#endif
+
+void __trigger_all_cpu_backtrace(void)
+{
+	int i;
+
+	backtrace_mask = cpu_online_map;
+	/* Wait for up to 10 seconds for all CPUs to do the backtrace */
+	for (i = 0; i < 10 * 1000; i++) {
+		if (cpus_empty(backtrace_mask))
+			break;
+		mdelay(1);
+	}
+}
+
+EXPORT_SYMBOL(nmi_active);
+EXPORT_SYMBOL(nmi_watchdog);
+EXPORT_SYMBOL(touch_nmi_watchdog);

arch/x86/kernel/pci-dma_64.c

@@ -0,0 +1,346 @@
+/*
+ * Dynamic DMA mapping support.
+ */
+
+#include <linux/types.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/pci.h>
+#include <linux/module.h>
+#include <asm/io.h>
+#include <asm/iommu.h>
+#include <asm/calgary.h>
+
+int iommu_merge __read_mostly = 0;
+EXPORT_SYMBOL(iommu_merge);
+
+dma_addr_t bad_dma_address __read_mostly;
+EXPORT_SYMBOL(bad_dma_address);
+
+/* This tells the BIO block layer to assume merging. Default to off
+   because we cannot guarantee merging later. */
+int iommu_bio_merge __read_mostly = 0;
+EXPORT_SYMBOL(iommu_bio_merge);
+
+static int iommu_sac_force __read_mostly = 0;
+
+int no_iommu __read_mostly;
+#ifdef CONFIG_IOMMU_DEBUG
+int panic_on_overflow __read_mostly = 1;
+int force_iommu __read_mostly = 1;
+#else
+int panic_on_overflow __read_mostly = 0;
+int force_iommu __read_mostly= 0;
+#endif
+
+/* Set this to 1 if there is a HW IOMMU in the system */
+int iommu_detected __read_mostly = 0;
+
+/* Dummy device used for NULL arguments (normally ISA). Better would
+   be probably a smaller DMA mask, but this is bug-to-bug compatible
+   to i386. */
+struct device fallback_dev = {
+	.bus_id = "fallback device",
+	.coherent_dma_mask = DMA_32BIT_MASK,
+	.dma_mask = &fallback_dev.coherent_dma_mask,
+};
+
+/* Allocate DMA memory on node near device */
+noinline static void *
+dma_alloc_pages(struct device *dev, gfp_t gfp, unsigned order)
+{
+	struct page *page;
+	int node;
+#ifdef CONFIG_PCI
+	if (dev->bus == &pci_bus_type)
+		node = pcibus_to_node(to_pci_dev(dev)->bus);
+	else
+#endif
+		node = numa_node_id();
+
+	if (node < first_node(node_online_map))
+		node = first_node(node_online_map);
+
+	page = alloc_pages_node(node, gfp, order);
+	return page ? page_address(page) : NULL;
+}
+
+/*
+ * Allocate memory for a coherent mapping.
+ */
+void *
+dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle,
+		   gfp_t gfp)
+{
+	void *memory;
+	unsigned long dma_mask = 0;
+	u64 bus;
+
+	if (!dev)
+		dev = &fallback_dev;
+	dma_mask = dev->coherent_dma_mask;
+	if (dma_mask == 0)
+		dma_mask = DMA_32BIT_MASK;
+
+	/* Device not DMA able */
+	if (dev->dma_mask == NULL)
+		return NULL;
+
+	/* Don't invoke OOM killer */
+	gfp |= __GFP_NORETRY;
+
+	/* Kludge to make it bug-to-bug compatible with i386. i386
+	   uses the normal dma_mask for alloc_coherent. */
+	dma_mask &= *dev->dma_mask;
+
+	/* Why <=? Even when the mask is smaller than 4GB it is often
+	   larger than 16MB and in this case we have a chance of
+	   finding fitting memory in the next higher zone first. If
+	   not retry with true GFP_DMA. -AK */
+	if (dma_mask <= DMA_32BIT_MASK)
+		gfp |= GFP_DMA32;
+
+ again:
+	memory = dma_alloc_pages(dev, gfp, get_order(size));
+	if (memory == NULL)
+		return NULL;
+
+	{
+		int high, mmu;
+		bus = virt_to_bus(memory);
+	        high = (bus + size) >= dma_mask;
+		mmu = high;
+		if (force_iommu && !(gfp & GFP_DMA))
+			mmu = 1;
+		else if (high) {
+			free_pages((unsigned long)memory,
+				   get_order(size));
+
+			/* Don't use the 16MB ZONE_DMA unless absolutely
+			   needed. It's better to use remapping first. */
+			if (dma_mask < DMA_32BIT_MASK && !(gfp & GFP_DMA)) {
+				gfp = (gfp & ~GFP_DMA32) | GFP_DMA;
+				goto again;
+			}
+
+			/* Let low level make its own zone decisions */
+			gfp &= ~(GFP_DMA32|GFP_DMA);
+
+			if (dma_ops->alloc_coherent)
+				return dma_ops->alloc_coherent(dev, size,
+							   dma_handle, gfp);
+			return NULL;
+		}
+
+		memset(memory, 0, size);
+		if (!mmu) {
+			*dma_handle = virt_to_bus(memory);
+			return memory;
+		}
+	}
+
+	if (dma_ops->alloc_coherent) {
+		free_pages((unsigned long)memory, get_order(size));
+		gfp &= ~(GFP_DMA|GFP_DMA32);
+		return dma_ops->alloc_coherent(dev, size, dma_handle, gfp);
+	}
+
+	if (dma_ops->map_simple) {
+		*dma_handle = dma_ops->map_simple(dev, memory,
+					      size,
+					      PCI_DMA_BIDIRECTIONAL);
+		if (*dma_handle != bad_dma_address)
+			return memory;
+	}
+
+	if (panic_on_overflow)
+		panic("dma_alloc_coherent: IOMMU overflow by %lu bytes\n",size);
+	free_pages((unsigned long)memory, get_order(size));
+	return NULL;
+}
+EXPORT_SYMBOL(dma_alloc_coherent);
+
+/*
+ * Unmap coherent memory.
+ * The caller must ensure that the device has finished accessing the mapping.
+ */
+void dma_free_coherent(struct device *dev, size_t size,
+			 void *vaddr, dma_addr_t bus)
+{
+	if (dma_ops->unmap_single)
+		dma_ops->unmap_single(dev, bus, size, 0);
+	free_pages((unsigned long)vaddr, get_order(size));
+}
+EXPORT_SYMBOL(dma_free_coherent);
+
+static int forbid_dac __read_mostly;
+
+int dma_supported(struct device *dev, u64 mask)
+{
+#ifdef CONFIG_PCI
+	if (mask > 0xffffffff && forbid_dac > 0) {
+
+
+
+		printk(KERN_INFO "PCI: Disallowing DAC for device %s\n", dev->bus_id);
+		return 0;
+	}
+#endif
+
+	if (dma_ops->dma_supported)
+		return dma_ops->dma_supported(dev, mask);
+
+	/* Copied from i386. Doesn't make much sense, because it will
+	   only work for pci_alloc_coherent.
+	   The caller just has to use GFP_DMA in this case. */
+        if (mask < DMA_24BIT_MASK)
+                return 0;
+
+	/* Tell the device to use SAC when IOMMU force is on.  This
+	   allows the driver to use cheaper accesses in some cases.
+
+	   Problem with this is that if we overflow the IOMMU area and
+	   return DAC as fallback address the device may not handle it
+	   correctly.
+
+	   As a special case some controllers have a 39bit address
+	   mode that is as efficient as 32bit (aic79xx). Don't force
+	   SAC for these.  Assume all masks <= 40 bits are of this
+	   type. Normally this doesn't make any difference, but gives
+	   more gentle handling of IOMMU overflow. */
+	if (iommu_sac_force && (mask >= DMA_40BIT_MASK)) {
+		printk(KERN_INFO "%s: Force SAC with mask %Lx\n", dev->bus_id,mask);
+		return 0;
+	}
+
+	return 1;
+}
+EXPORT_SYMBOL(dma_supported);
+
+int dma_set_mask(struct device *dev, u64 mask)
+{
+	if (!dev->dma_mask || !dma_supported(dev, mask))
+		return -EIO;
+	*dev->dma_mask = mask;
+	return 0;
+}
+EXPORT_SYMBOL(dma_set_mask);
+
+/*
+ * See <Documentation/x86_64/boot-options.txt> for the iommu kernel parameter
+ * documentation.
+ */
+__init int iommu_setup(char *p)
+{
+	iommu_merge = 1;
+
+	if (!p)
+		return -EINVAL;
+
+	while (*p) {
+		if (!strncmp(p,"off",3))
+			no_iommu = 1;
+		/* gart_parse_options has more force support */
+		if (!strncmp(p,"force",5))
+			force_iommu = 1;
+		if (!strncmp(p,"noforce",7)) {
+			iommu_merge = 0;
+			force_iommu = 0;
+		}
+
+		if (!strncmp(p, "biomerge",8)) {
+			iommu_bio_merge = 4096;
+			iommu_merge = 1;
+			force_iommu = 1;
+		}
+		if (!strncmp(p, "panic",5))
+			panic_on_overflow = 1;
+		if (!strncmp(p, "nopanic",7))
+			panic_on_overflow = 0;
+		if (!strncmp(p, "merge",5)) {
+			iommu_merge = 1;
+			force_iommu = 1;
+		}
+		if (!strncmp(p, "nomerge",7))
+			iommu_merge = 0;
+		if (!strncmp(p, "forcesac",8))
+			iommu_sac_force = 1;
+		if (!strncmp(p, "allowdac", 8))
+			forbid_dac = 0;
+		if (!strncmp(p, "nodac", 5))
+			forbid_dac = -1;
+
+#ifdef CONFIG_SWIOTLB
+		if (!strncmp(p, "soft",4))
+			swiotlb = 1;
+#endif
+
+#ifdef CONFIG_IOMMU
+		gart_parse_options(p);
+#endif
+
+#ifdef CONFIG_CALGARY_IOMMU
+		if (!strncmp(p, "calgary", 7))
+			use_calgary = 1;
+#endif /* CONFIG_CALGARY_IOMMU */
+
+		p += strcspn(p, ",");
+		if (*p == ',')
+			++p;
+	}
+	return 0;
+}
+early_param("iommu", iommu_setup);
+
+void __init pci_iommu_alloc(void)
+{
+	/*
+	 * The order of these functions is important for
+	 * fall-back/fail-over reasons
+	 */
+#ifdef CONFIG_IOMMU
+	iommu_hole_init();
+#endif
+
+#ifdef CONFIG_CALGARY_IOMMU
+	detect_calgary();
+#endif
+
+#ifdef CONFIG_SWIOTLB
+	pci_swiotlb_init();
+#endif
+}
+
+static int __init pci_iommu_init(void)
+{
+#ifdef CONFIG_CALGARY_IOMMU
+	calgary_iommu_init();
+#endif
+
+#ifdef CONFIG_IOMMU
+	gart_iommu_init();
+#endif
+
+	no_iommu_init();
+	return 0;
+}
+
+void pci_iommu_shutdown(void)
+{
+	gart_iommu_shutdown();
+}
+
+#ifdef CONFIG_PCI
+/* Many VIA bridges seem to corrupt data for DAC. Disable it here */
+
+static __devinit void via_no_dac(struct pci_dev *dev)
+{
+	if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI && forbid_dac == 0) {
+		printk(KERN_INFO "PCI: VIA PCI bridge detected. Disabling DAC.\n");
+		forbid_dac = 1;
+	}
+}
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_ANY_ID, via_no_dac);
+#endif
+/* Must execute after PCI subsystem */
+fs_initcall(pci_iommu_init);

arch/x86/kernel/pci-gart_64.c

@@ -0,0 +1,740 @@
+/*
+ * Dynamic DMA mapping support for AMD Hammer.
+ * 
+ * Use the integrated AGP GART in the Hammer northbridge as an IOMMU for PCI.
+ * This allows to use PCI devices that only support 32bit addresses on systems
+ * with more than 4GB. 
+ *
+ * See Documentation/DMA-mapping.txt for the interface specification.
+ * 
+ * Copyright 2002 Andi Kleen, SuSE Labs.
+ */
+
+#include <linux/types.h>
+#include <linux/ctype.h>
+#include <linux/agp_backend.h>
+#include <linux/init.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/spinlock.h>
+#include <linux/pci.h>
+#include <linux/module.h>
+#include <linux/topology.h>
+#include <linux/interrupt.h>
+#include <linux/bitops.h>
+#include <linux/kdebug.h>
+#include <asm/atomic.h>
+#include <asm/io.h>
+#include <asm/mtrr.h>
+#include <asm/pgtable.h>
+#include <asm/proto.h>
+#include <asm/iommu.h>
+#include <asm/cacheflush.h>
+#include <asm/swiotlb.h>
+#include <asm/dma.h>
+#include <asm/k8.h>
+
+unsigned long iommu_bus_base;	/* GART remapping area (physical) */
+static unsigned long iommu_size; 	/* size of remapping area bytes */
+static unsigned long iommu_pages;	/* .. and in pages */
+
+u32 *iommu_gatt_base; 		/* Remapping table */
+
+/* If this is disabled the IOMMU will use an optimized flushing strategy
+   of only flushing when an mapping is reused. With it true the GART is flushed 
+   for every mapping. Problem is that doing the lazy flush seems to trigger
+   bugs with some popular PCI cards, in particular 3ware (but has been also
+   also seen with Qlogic at least). */
+int iommu_fullflush = 1;
+
+/* Allocation bitmap for the remapping area */ 
+static DEFINE_SPINLOCK(iommu_bitmap_lock);
+static unsigned long *iommu_gart_bitmap; /* guarded by iommu_bitmap_lock */
+
+static u32 gart_unmapped_entry; 
+
+#define GPTE_VALID    1
+#define GPTE_COHERENT 2
+#define GPTE_ENCODE(x) \
+	(((x) & 0xfffff000) | (((x) >> 32) << 4) | GPTE_VALID | GPTE_COHERENT)
+#define GPTE_DECODE(x) (((x) & 0xfffff000) | (((u64)(x) & 0xff0) << 28))
+
+#define to_pages(addr,size) \
+	(round_up(((addr) & ~PAGE_MASK) + (size), PAGE_SIZE) >> PAGE_SHIFT)
+
+#define EMERGENCY_PAGES 32 /* = 128KB */ 
+
+#ifdef CONFIG_AGP
+#define AGPEXTERN extern
+#else
+#define AGPEXTERN
+#endif
+
+/* backdoor interface to AGP driver */
+AGPEXTERN int agp_memory_reserved;
+AGPEXTERN __u32 *agp_gatt_table;
+
+static unsigned long next_bit;  /* protected by iommu_bitmap_lock */
+static int need_flush; 		/* global flush state. set for each gart wrap */
+
+static unsigned long alloc_iommu(int size) 
+{ 	
+	unsigned long offset, flags;
+
+	spin_lock_irqsave(&iommu_bitmap_lock, flags);	
+	offset = find_next_zero_string(iommu_gart_bitmap,next_bit,iommu_pages,size);
+	if (offset == -1) {
+		need_flush = 1;
+		offset = find_next_zero_string(iommu_gart_bitmap,0,iommu_pages,size);
+	}
+	if (offset != -1) { 
+		set_bit_string(iommu_gart_bitmap, offset, size); 
+		next_bit = offset+size; 
+		if (next_bit >= iommu_pages) { 
+			next_bit = 0;
+			need_flush = 1;
+		} 
+	} 
+	if (iommu_fullflush)
+		need_flush = 1;
+	spin_unlock_irqrestore(&iommu_bitmap_lock, flags);      
+	return offset;
+} 
+
+static void free_iommu(unsigned long offset, int size)
+{ 
+	unsigned long flags;
+	spin_lock_irqsave(&iommu_bitmap_lock, flags);
+	__clear_bit_string(iommu_gart_bitmap, offset, size);
+	spin_unlock_irqrestore(&iommu_bitmap_lock, flags);
+} 
+
+/* 
+ * Use global flush state to avoid races with multiple flushers.
+ */
+static void flush_gart(void)
+{ 
+	unsigned long flags;
+	spin_lock_irqsave(&iommu_bitmap_lock, flags);
+	if (need_flush) {
+		k8_flush_garts();
+		need_flush = 0;
+	} 
+	spin_unlock_irqrestore(&iommu_bitmap_lock, flags);
+} 
+
+#ifdef CONFIG_IOMMU_LEAK
+
+#define SET_LEAK(x) if (iommu_leak_tab) \
+			iommu_leak_tab[x] = __builtin_return_address(0);
+#define CLEAR_LEAK(x) if (iommu_leak_tab) \
+			iommu_leak_tab[x] = NULL;
+
+/* Debugging aid for drivers that don't free their IOMMU tables */
+static void **iommu_leak_tab; 
+static int leak_trace;
+int iommu_leak_pages = 20; 
+void dump_leak(void)
+{
+	int i;
+	static int dump; 
+	if (dump || !iommu_leak_tab) return;
+	dump = 1;
+	show_stack(NULL,NULL);
+	/* Very crude. dump some from the end of the table too */ 
+	printk("Dumping %d pages from end of IOMMU:\n", iommu_leak_pages); 
+	for (i = 0; i < iommu_leak_pages; i+=2) {
+		printk("%lu: ", iommu_pages-i);
+		printk_address((unsigned long) iommu_leak_tab[iommu_pages-i]);
+		printk("%c", (i+1)%2 == 0 ? '\n' : ' '); 
+	} 
+	printk("\n");
+}
+#else
+#define SET_LEAK(x)
+#define CLEAR_LEAK(x)
+#endif
+
+static void iommu_full(struct device *dev, size_t size, int dir)
+{
+	/* 
+	 * Ran out of IOMMU space for this operation. This is very bad.
+	 * Unfortunately the drivers cannot handle this operation properly.
+	 * Return some non mapped prereserved space in the aperture and 
+	 * let the Northbridge deal with it. This will result in garbage
+	 * in the IO operation. When the size exceeds the prereserved space
+	 * memory corruption will occur or random memory will be DMAed 
+	 * out. Hopefully no network devices use single mappings that big.
+	 */ 
+	
+	printk(KERN_ERR 
+  "PCI-DMA: Out of IOMMU space for %lu bytes at device %s\n",
+	       size, dev->bus_id);
+
+	if (size > PAGE_SIZE*EMERGENCY_PAGES) {
+		if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
+			panic("PCI-DMA: Memory would be corrupted\n");
+		if (dir == PCI_DMA_TODEVICE || dir == PCI_DMA_BIDIRECTIONAL) 
+			panic(KERN_ERR "PCI-DMA: Random memory would be DMAed\n");
+	} 
+
+#ifdef CONFIG_IOMMU_LEAK
+	dump_leak(); 
+#endif
+} 
+
+static inline int need_iommu(struct device *dev, unsigned long addr, size_t size)
+{ 
+	u64 mask = *dev->dma_mask;
+	int high = addr + size > mask;
+	int mmu = high;
+	if (force_iommu) 
+		mmu = 1; 
+	return mmu; 
+}
+
+static inline int nonforced_iommu(struct device *dev, unsigned long addr, size_t size)
+{ 
+	u64 mask = *dev->dma_mask;
+	int high = addr + size > mask;
+	int mmu = high;
+	return mmu; 
+}
+
+/* Map a single continuous physical area into the IOMMU.
+ * Caller needs to check if the iommu is needed and flush.
+ */
+static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem,
+				size_t size, int dir)
+{ 
+	unsigned long npages = to_pages(phys_mem, size);
+	unsigned long iommu_page = alloc_iommu(npages);
+	int i;
+	if (iommu_page == -1) {
+		if (!nonforced_iommu(dev, phys_mem, size))
+			return phys_mem; 
+		if (panic_on_overflow)
+			panic("dma_map_area overflow %lu bytes\n", size);
+		iommu_full(dev, size, dir);
+		return bad_dma_address;
+	}
+
+	for (i = 0; i < npages; i++) {
+		iommu_gatt_base[iommu_page + i] = GPTE_ENCODE(phys_mem);
+		SET_LEAK(iommu_page + i);
+		phys_mem += PAGE_SIZE;
+	}
+	return iommu_bus_base + iommu_page*PAGE_SIZE + (phys_mem & ~PAGE_MASK);
+}
+
+static dma_addr_t gart_map_simple(struct device *dev, char *buf,
+				 size_t size, int dir)
+{
+	dma_addr_t map = dma_map_area(dev, virt_to_bus(buf), size, dir);
+	flush_gart();
+	return map;
+}
+
+/* Map a single area into the IOMMU */
+static dma_addr_t gart_map_single(struct device *dev, void *addr, size_t size, int dir)
+{
+	unsigned long phys_mem, bus;
+
+	if (!dev)
+		dev = &fallback_dev;
+
+	phys_mem = virt_to_phys(addr); 
+	if (!need_iommu(dev, phys_mem, size))
+		return phys_mem; 
+
+	bus = gart_map_simple(dev, addr, size, dir);
+	return bus; 
+}
+
+/*
+ * Free a DMA mapping.
+ */
+static void gart_unmap_single(struct device *dev, dma_addr_t dma_addr,
+		      size_t size, int direction)
+{
+	unsigned long iommu_page;
+	int npages;
+	int i;
+
+	if (dma_addr < iommu_bus_base + EMERGENCY_PAGES*PAGE_SIZE ||
+	    dma_addr >= iommu_bus_base + iommu_size)
+		return;
+	iommu_page = (dma_addr - iommu_bus_base)>>PAGE_SHIFT;
+	npages = to_pages(dma_addr, size);
+	for (i = 0; i < npages; i++) {
+		iommu_gatt_base[iommu_page + i] = gart_unmapped_entry;
+		CLEAR_LEAK(iommu_page + i);
+	}
+	free_iommu(iommu_page, npages);
+}
+
+/*
+ * Wrapper for pci_unmap_single working with scatterlists.
+ */
+static void gart_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, int dir)
+{
+	int i;
+
+	for (i = 0; i < nents; i++) {
+		struct scatterlist *s = &sg[i];
+		if (!s->dma_length || !s->length)
+			break;
+		gart_unmap_single(dev, s->dma_address, s->dma_length, dir);
+	}
+}
+
+/* Fallback for dma_map_sg in case of overflow */
+static int dma_map_sg_nonforce(struct device *dev, struct scatterlist *sg,
+			       int nents, int dir)
+{
+	int i;
+
+#ifdef CONFIG_IOMMU_DEBUG
+	printk(KERN_DEBUG "dma_map_sg overflow\n");
+#endif
+
+ 	for (i = 0; i < nents; i++ ) {
+		struct scatterlist *s = &sg[i];
+		unsigned long addr = page_to_phys(s->page) + s->offset; 
+		if (nonforced_iommu(dev, addr, s->length)) { 
+			addr = dma_map_area(dev, addr, s->length, dir);
+			if (addr == bad_dma_address) { 
+				if (i > 0) 
+					gart_unmap_sg(dev, sg, i, dir);
+				nents = 0; 
+				sg[0].dma_length = 0;
+				break;
+			}
+		}
+		s->dma_address = addr;
+		s->dma_length = s->length;
+	}
+	flush_gart();
+	return nents;
+}
+
+/* Map multiple scatterlist entries continuous into the first. */
+static int __dma_map_cont(struct scatterlist *sg, int start, int stopat,
+		      struct scatterlist *sout, unsigned long pages)
+{
+	unsigned long iommu_start = alloc_iommu(pages);
+	unsigned long iommu_page = iommu_start; 
+	int i;
+
+	if (iommu_start == -1)
+		return -1;
+	
+	for (i = start; i < stopat; i++) {
+		struct scatterlist *s = &sg[i];
+		unsigned long pages, addr;
+		unsigned long phys_addr = s->dma_address;
+		
+		BUG_ON(i > start && s->offset);
+		if (i == start) {
+			*sout = *s; 
+			sout->dma_address = iommu_bus_base;
+			sout->dma_address += iommu_page*PAGE_SIZE + s->offset;
+			sout->dma_length = s->length;
+		} else { 
+			sout->dma_length += s->length; 
+		}
+
+		addr = phys_addr;
+		pages = to_pages(s->offset, s->length); 
+		while (pages--) { 
+			iommu_gatt_base[iommu_page] = GPTE_ENCODE(addr); 
+			SET_LEAK(iommu_page);
+			addr += PAGE_SIZE;
+			iommu_page++;
+		}
+	} 
+	BUG_ON(iommu_page - iommu_start != pages);	
+	return 0;
+}
+
+static inline int dma_map_cont(struct scatterlist *sg, int start, int stopat,
+		      struct scatterlist *sout,
+		      unsigned long pages, int need)
+{
+	if (!need) { 
+		BUG_ON(stopat - start != 1);
+		*sout = sg[start]; 
+		sout->dma_length = sg[start].length; 
+		return 0;
+	} 
+	return __dma_map_cont(sg, start, stopat, sout, pages);
+}
+		
+/*
+ * DMA map all entries in a scatterlist.
+ * Merge chunks that have page aligned sizes into a continuous mapping. 
+ */
+int gart_map_sg(struct device *dev, struct scatterlist *sg, int nents, int dir)
+{
+	int i;
+	int out;
+	int start;
+	unsigned long pages = 0;
+	int need = 0, nextneed;
+
+	if (nents == 0) 
+		return 0;
+
+	if (!dev)
+		dev = &fallback_dev;
+
+	out = 0;
+	start = 0;
+	for (i = 0; i < nents; i++) {
+		struct scatterlist *s = &sg[i];
+		dma_addr_t addr = page_to_phys(s->page) + s->offset;
+		s->dma_address = addr;
+		BUG_ON(s->length == 0); 
+
+		nextneed = need_iommu(dev, addr, s->length); 
+
+		/* Handle the previous not yet processed entries */
+		if (i > start) {
+			struct scatterlist *ps = &sg[i-1];
+			/* Can only merge when the last chunk ends on a page 
+			   boundary and the new one doesn't have an offset. */
+			if (!iommu_merge || !nextneed || !need || s->offset ||
+			    (ps->offset + ps->length) % PAGE_SIZE) { 
+				if (dma_map_cont(sg, start, i, sg+out, pages,
+						 need) < 0)
+					goto error;
+				out++;
+				pages = 0;
+				start = i;	
+			}
+		}
+
+		need = nextneed;
+		pages += to_pages(s->offset, s->length);
+	}
+	if (dma_map_cont(sg, start, i, sg+out, pages, need) < 0)
+		goto error;
+	out++;
+	flush_gart();
+	if (out < nents) 
+		sg[out].dma_length = 0; 
+	return out;
+
+error:
+	flush_gart();
+	gart_unmap_sg(dev, sg, nents, dir);
+	/* When it was forced or merged try again in a dumb way */
+	if (force_iommu || iommu_merge) {
+		out = dma_map_sg_nonforce(dev, sg, nents, dir);
+		if (out > 0)
+			return out;
+	}
+	if (panic_on_overflow)
+		panic("dma_map_sg: overflow on %lu pages\n", pages);
+	iommu_full(dev, pages << PAGE_SHIFT, dir);
+	for (i = 0; i < nents; i++)
+		sg[i].dma_address = bad_dma_address;
+	return 0;
+} 
+
+static int no_agp;
+
+static __init unsigned long check_iommu_size(unsigned long aper, u64 aper_size)
+{ 
+	unsigned long a; 
+	if (!iommu_size) { 
+		iommu_size = aper_size; 
+		if (!no_agp) 
+			iommu_size /= 2; 
+	} 
+
+	a = aper + iommu_size; 
+	iommu_size -= round_up(a, LARGE_PAGE_SIZE) - a;
+
+	if (iommu_size < 64*1024*1024) 
+		printk(KERN_WARNING
+  "PCI-DMA: Warning: Small IOMMU %luMB. Consider increasing the AGP aperture in BIOS\n",iommu_size>>20); 
+	
+	return iommu_size;
+} 
+
+static __init unsigned read_aperture(struct pci_dev *dev, u32 *size) 
+{ 
+	unsigned aper_size = 0, aper_base_32;
+	u64 aper_base;
+	unsigned aper_order;
+
+	pci_read_config_dword(dev, 0x94, &aper_base_32); 
+	pci_read_config_dword(dev, 0x90, &aper_order);
+	aper_order = (aper_order >> 1) & 7;	
+
+	aper_base = aper_base_32 & 0x7fff; 
+	aper_base <<= 25;
+
+	aper_size = (32 * 1024 * 1024) << aper_order; 
+       if (aper_base + aper_size > 0x100000000UL || !aper_size)
+		aper_base = 0;
+
+	*size = aper_size;
+	return aper_base;
+} 
+
+/* 
+ * Private Northbridge GATT initialization in case we cannot use the
+ * AGP driver for some reason.  
+ */
+static __init int init_k8_gatt(struct agp_kern_info *info)
+{ 
+	struct pci_dev *dev;
+	void *gatt;
+	unsigned aper_base, new_aper_base;
+	unsigned aper_size, gatt_size, new_aper_size;
+	int i;
+
+	printk(KERN_INFO "PCI-DMA: Disabling AGP.\n");
+	aper_size = aper_base = info->aper_size = 0;
+	dev = NULL;
+	for (i = 0; i < num_k8_northbridges; i++) {
+		dev = k8_northbridges[i];
+		new_aper_base = read_aperture(dev, &new_aper_size); 
+		if (!new_aper_base) 
+			goto nommu; 
+		
+		if (!aper_base) { 
+			aper_size = new_aper_size;
+			aper_base = new_aper_base;
+		} 
+		if (aper_size != new_aper_size || aper_base != new_aper_base) 
+			goto nommu;
+	}
+	if (!aper_base)
+		goto nommu; 
+	info->aper_base = aper_base;
+	info->aper_size = aper_size>>20; 
+
+	gatt_size = (aper_size >> PAGE_SHIFT) * sizeof(u32); 
+	gatt = (void *)__get_free_pages(GFP_KERNEL, get_order(gatt_size)); 
+	if (!gatt) 
+		panic("Cannot allocate GATT table");
+	if (change_page_attr_addr((unsigned long)gatt, gatt_size >> PAGE_SHIFT, PAGE_KERNEL_NOCACHE))
+		panic("Could not set GART PTEs to uncacheable pages");
+	global_flush_tlb();
+
+	memset(gatt, 0, gatt_size); 
+	agp_gatt_table = gatt;
+
+	for (i = 0; i < num_k8_northbridges; i++) {
+		u32 ctl; 
+		u32 gatt_reg; 
+
+		dev = k8_northbridges[i];
+		gatt_reg = __pa(gatt) >> 12; 
+		gatt_reg <<= 4; 
+		pci_write_config_dword(dev, 0x98, gatt_reg);
+		pci_read_config_dword(dev, 0x90, &ctl); 
+
+		ctl |= 1;
+		ctl &= ~((1<<4) | (1<<5));
+
+		pci_write_config_dword(dev, 0x90, ctl); 
+	}
+	flush_gart();
+	
+	printk("PCI-DMA: aperture base @ %x size %u KB\n",aper_base, aper_size>>10); 
+	return 0;
+
+ nommu:
+ 	/* Should not happen anymore */
+	printk(KERN_ERR "PCI-DMA: More than 4GB of RAM and no IOMMU\n"
+	       KERN_ERR "PCI-DMA: 32bit PCI IO may malfunction.\n");
+	return -1; 
+} 
+
+extern int agp_amd64_init(void);
+
+static const struct dma_mapping_ops gart_dma_ops = {
+	.mapping_error = NULL,
+	.map_single = gart_map_single,
+	.map_simple = gart_map_simple,
+	.unmap_single = gart_unmap_single,
+	.sync_single_for_cpu = NULL,
+	.sync_single_for_device = NULL,
+	.sync_single_range_for_cpu = NULL,
+	.sync_single_range_for_device = NULL,
+	.sync_sg_for_cpu = NULL,
+	.sync_sg_for_device = NULL,
+	.map_sg = gart_map_sg,
+	.unmap_sg = gart_unmap_sg,
+};
+
+void gart_iommu_shutdown(void)
+{
+	struct pci_dev *dev;
+	int i;
+
+	if (no_agp && (dma_ops != &gart_dma_ops))
+		return;
+
+        for (i = 0; i < num_k8_northbridges; i++) {
+                u32 ctl;
+
+                dev = k8_northbridges[i];
+                pci_read_config_dword(dev, 0x90, &ctl);
+
+                ctl &= ~1;
+
+                pci_write_config_dword(dev, 0x90, ctl);
+        }
+}
+
+void __init gart_iommu_init(void)
+{ 
+	struct agp_kern_info info;
+	unsigned long aper_size;
+	unsigned long iommu_start;
+	unsigned long scratch;
+	long i;
+
+	if (cache_k8_northbridges() < 0 || num_k8_northbridges == 0) {
+		printk(KERN_INFO "PCI-GART: No AMD northbridge found.\n");
+		return;
+	}
+
+#ifndef CONFIG_AGP_AMD64
+	no_agp = 1; 
+#else
+	/* Makefile puts PCI initialization via subsys_initcall first. */
+	/* Add other K8 AGP bridge drivers here */
+	no_agp = no_agp || 
+		(agp_amd64_init() < 0) || 
+		(agp_copy_info(agp_bridge, &info) < 0);
+#endif	
+
+	if (swiotlb)
+		return;
+
+	/* Did we detect a different HW IOMMU? */
+	if (iommu_detected && !iommu_aperture)
+		return;
+
+	if (no_iommu ||
+	    (!force_iommu && end_pfn <= MAX_DMA32_PFN) ||
+	    !iommu_aperture ||
+	    (no_agp && init_k8_gatt(&info) < 0)) {
+		if (end_pfn > MAX_DMA32_PFN) {
+			printk(KERN_ERR "WARNING more than 4GB of memory "
+					"but GART IOMMU not available.\n"
+			       KERN_ERR "WARNING 32bit PCI may malfunction.\n");
+		}
+		return;
+	}
+
+	printk(KERN_INFO "PCI-DMA: using GART IOMMU.\n");
+	aper_size = info.aper_size * 1024 * 1024;	
+	iommu_size = check_iommu_size(info.aper_base, aper_size); 
+	iommu_pages = iommu_size >> PAGE_SHIFT; 
+
+	iommu_gart_bitmap = (void*)__get_free_pages(GFP_KERNEL, 
+						    get_order(iommu_pages/8)); 
+	if (!iommu_gart_bitmap) 
+		panic("Cannot allocate iommu bitmap\n"); 
+	memset(iommu_gart_bitmap, 0, iommu_pages/8);
+
+#ifdef CONFIG_IOMMU_LEAK
+	if (leak_trace) { 
+		iommu_leak_tab = (void *)__get_free_pages(GFP_KERNEL, 
+				  get_order(iommu_pages*sizeof(void *)));
+		if (iommu_leak_tab) 
+			memset(iommu_leak_tab, 0, iommu_pages * 8); 
+		else
+			printk("PCI-DMA: Cannot allocate leak trace area\n"); 
+	} 
+#endif
+
+	/* 
+	 * Out of IOMMU space handling.
+	 * Reserve some invalid pages at the beginning of the GART. 
+	 */ 
+	set_bit_string(iommu_gart_bitmap, 0, EMERGENCY_PAGES); 
+
+	agp_memory_reserved = iommu_size;	
+	printk(KERN_INFO
+	       "PCI-DMA: Reserving %luMB of IOMMU area in the AGP aperture\n",
+	       iommu_size>>20); 
+
+	iommu_start = aper_size - iommu_size;	
+	iommu_bus_base = info.aper_base + iommu_start; 
+	bad_dma_address = iommu_bus_base;
+	iommu_gatt_base = agp_gatt_table + (iommu_start>>PAGE_SHIFT);
+
+	/* 
+	 * Unmap the IOMMU part of the GART. The alias of the page is
+	 * always mapped with cache enabled and there is no full cache
+	 * coherency across the GART remapping. The unmapping avoids
+	 * automatic prefetches from the CPU allocating cache lines in
+	 * there. All CPU accesses are done via the direct mapping to
+	 * the backing memory. The GART address is only used by PCI
+	 * devices. 
+	 */
+	clear_kernel_mapping((unsigned long)__va(iommu_bus_base), iommu_size);
+
+	/* 
+	 * Try to workaround a bug (thanks to BenH) 
+	 * Set unmapped entries to a scratch page instead of 0. 
+	 * Any prefetches that hit unmapped entries won't get an bus abort
+	 * then.
+	 */
+	scratch = get_zeroed_page(GFP_KERNEL); 
+	if (!scratch) 
+		panic("Cannot allocate iommu scratch page");
+	gart_unmapped_entry = GPTE_ENCODE(__pa(scratch));
+	for (i = EMERGENCY_PAGES; i < iommu_pages; i++) 
+		iommu_gatt_base[i] = gart_unmapped_entry;
+
+	flush_gart();
+	dma_ops = &gart_dma_ops;
+} 
+
+void __init gart_parse_options(char *p)
+{
+	int arg;
+
+#ifdef CONFIG_IOMMU_LEAK
+	if (!strncmp(p,"leak",4)) {
+		leak_trace = 1;
+		p += 4;
+		if (*p == '=') ++p;
+		if (isdigit(*p) && get_option(&p, &arg))
+			iommu_leak_pages = arg;
+	}
+#endif
+	if (isdigit(*p) && get_option(&p, &arg))
+		iommu_size = arg;
+	if (!strncmp(p, "fullflush",8))
+		iommu_fullflush = 1;
+	if (!strncmp(p, "nofullflush",11))
+		iommu_fullflush = 0;
+	if (!strncmp(p,"noagp",5))
+		no_agp = 1;
+	if (!strncmp(p, "noaperture",10))
+		fix_aperture = 0;
+	/* duplicated from pci-dma.c */
+	if (!strncmp(p,"force",5))
+		iommu_aperture_allowed = 1;
+	if (!strncmp(p,"allowed",7))
+		iommu_aperture_allowed = 1;
+	if (!strncmp(p, "memaper", 7)) {
+		fallback_aper_force = 1;
+		p += 7;
+		if (*p == '=') {
+			++p;
+			if (get_option(&p, &arg))
+				fallback_aper_order = arg;
+		}
+	}
+}

arch/x86/kernel/pci-nommu_64.c

@@ -0,0 +1,97 @@
+/* Fallback functions when the main IOMMU code is not compiled in. This
+   code is roughly equivalent to i386. */
+#include <linux/mm.h>
+#include <linux/init.h>
+#include <linux/pci.h>
+#include <linux/string.h>
+#include <linux/dma-mapping.h>
+
+#include <asm/iommu.h>
+#include <asm/processor.h>
+#include <asm/dma.h>
+
+static int
+check_addr(char *name, struct device *hwdev, dma_addr_t bus, size_t size)
+{
+        if (hwdev && bus + size > *hwdev->dma_mask) {
+		if (*hwdev->dma_mask >= DMA_32BIT_MASK)
+			printk(KERN_ERR
+			    "nommu_%s: overflow %Lx+%zu of device mask %Lx\n",
+				name, (long long)bus, size,
+				(long long)*hwdev->dma_mask);
+		return 0;
+	}
+	return 1;
+}
+
+static dma_addr_t
+nommu_map_single(struct device *hwdev, void *ptr, size_t size,
+	       int direction)
+{
+	dma_addr_t bus = virt_to_bus(ptr);
+	if (!check_addr("map_single", hwdev, bus, size))
+				return bad_dma_address;
+	return bus;
+}
+
+static void nommu_unmap_single(struct device *dev, dma_addr_t addr,size_t size,
+			int direction)
+{
+}
+
+/* Map a set of buffers described by scatterlist in streaming
+ * mode for DMA.  This is the scatter-gather version of the
+ * above pci_map_single interface.  Here the scatter gather list
+ * elements are each tagged with the appropriate dma address
+ * and length.  They are obtained via sg_dma_{address,length}(SG).
+ *
+ * NOTE: An implementation may be able to use a smaller number of
+ *       DMA address/length pairs than there are SG table elements.
+ *       (for example via virtual mapping capabilities)
+ *       The routine returns the number of addr/length pairs actually
+ *       used, at most nents.
+ *
+ * Device ownership issues as mentioned above for pci_map_single are
+ * the same here.
+ */
+static int nommu_map_sg(struct device *hwdev, struct scatterlist *sg,
+	       int nents, int direction)
+{
+	int i;
+
+ 	for (i = 0; i < nents; i++ ) {
+		struct scatterlist *s = &sg[i];
+		BUG_ON(!s->page);
+		s->dma_address = virt_to_bus(page_address(s->page) +s->offset);
+		if (!check_addr("map_sg", hwdev, s->dma_address, s->length))
+			return 0;
+		s->dma_length = s->length;
+	}
+	return nents;
+}
+
+/* Unmap a set of streaming mode DMA translations.
+ * Again, cpu read rules concerning calls here are the same as for
+ * pci_unmap_single() above.
+ */
+static void nommu_unmap_sg(struct device *dev, struct scatterlist *sg,
+		  int nents, int dir)
+{
+}
+
+const struct dma_mapping_ops nommu_dma_ops = {
+	.map_single = nommu_map_single,
+	.unmap_single = nommu_unmap_single,
+	.map_sg = nommu_map_sg,
+	.unmap_sg = nommu_unmap_sg,
+	.is_phys = 1,
+};
+
+void __init no_iommu_init(void)
+{
+	if (dma_ops)
+		return;
+
+	force_iommu = 0; /* no HW IOMMU */
+	dma_ops = &nommu_dma_ops;
+}

arch/x86/kernel/pci-swiotlb_64.c

@@ -0,0 +1,44 @@
+/* Glue code to lib/swiotlb.c */
+
+#include <linux/pci.h>
+#include <linux/cache.h>
+#include <linux/module.h>
+#include <linux/dma-mapping.h>
+
+#include <asm/iommu.h>
+#include <asm/swiotlb.h>
+#include <asm/dma.h>
+
+int swiotlb __read_mostly;
+EXPORT_SYMBOL(swiotlb);
+
+const struct dma_mapping_ops swiotlb_dma_ops = {
+	.mapping_error = swiotlb_dma_mapping_error,
+	.alloc_coherent = swiotlb_alloc_coherent,
+	.free_coherent = swiotlb_free_coherent,
+	.map_single = swiotlb_map_single,
+	.unmap_single = swiotlb_unmap_single,
+	.sync_single_for_cpu = swiotlb_sync_single_for_cpu,
+	.sync_single_for_device = swiotlb_sync_single_for_device,
+	.sync_single_range_for_cpu = swiotlb_sync_single_range_for_cpu,
+	.sync_single_range_for_device = swiotlb_sync_single_range_for_device,
+	.sync_sg_for_cpu = swiotlb_sync_sg_for_cpu,
+	.sync_sg_for_device = swiotlb_sync_sg_for_device,
+	.map_sg = swiotlb_map_sg,
+	.unmap_sg = swiotlb_unmap_sg,
+	.dma_supported = NULL,
+};
+
+void __init pci_swiotlb_init(void)
+{
+	/* don't initialize swiotlb if iommu=off (no_iommu=1) */
+	if (!iommu_detected && !no_iommu && end_pfn > MAX_DMA32_PFN)
+	       swiotlb = 1;
+	if (swiotlb_force)
+		swiotlb = 1;
+	if (swiotlb) {
+		printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
+		swiotlb_init();
+		dma_ops = &swiotlb_dma_ops;
+	}
+}

arch/x86/kernel/pmtimer_64.c

@@ -0,0 +1,69 @@
+/* Ported over from i386 by AK, original copyright was:
+ *
+ * (C) Dominik Brodowski <linux@brodo.de> 2003
+ *
+ * Driver to use the Power Management Timer (PMTMR) available in some
+ * southbridges as primary timing source for the Linux kernel.
+ *
+ * Based on parts of linux/drivers/acpi/hardware/hwtimer.c, timer_pit.c,
+ * timer_hpet.c, and on Arjan van de Ven's implementation for 2.4.
+ *
+ * This file is licensed under the GPL v2.
+ *
+ * Dropped all the hardware bug workarounds for now. Hopefully they
+ * are not needed on 64bit chipsets.
+ */
+
+#include <linux/jiffies.h>
+#include <linux/kernel.h>
+#include <linux/time.h>
+#include <linux/init.h>
+#include <linux/cpumask.h>
+#include <asm/io.h>
+#include <asm/proto.h>
+#include <asm/msr.h>
+#include <asm/vsyscall.h>
+
+#define ACPI_PM_MASK 0xFFFFFF /* limit it to 24 bits */
+
+static inline u32 cyc2us(u32 cycles)
+{
+	/* The Power Management Timer ticks at 3.579545 ticks per microsecond.
+	 * 1 / PM_TIMER_FREQUENCY == 0.27936511 =~ 286/1024 [error: 0.024%]
+	 *
+	 * Even with HZ = 100, delta is at maximum 35796 ticks, so it can
+	 * easily be multiplied with 286 (=0x11E) without having to fear
+	 * u32 overflows.
+	 */
+	cycles *= 286;
+	return (cycles >> 10);
+}
+
+static unsigned pmtimer_wait_tick(void)
+{
+	u32 a, b;
+	for (a = b = inl(pmtmr_ioport) & ACPI_PM_MASK;
+	     a == b;
+	     b = inl(pmtmr_ioport) & ACPI_PM_MASK)
+		cpu_relax();
+	return b;
+}
+
+/* note: wait time is rounded up to one tick */
+void pmtimer_wait(unsigned us)
+{
+	u32 a, b;
+	a = pmtimer_wait_tick();
+	do {
+		b = inl(pmtmr_ioport);
+		cpu_relax();
+	} while (cyc2us(b - a) < us);
+}
+
+static int __init nopmtimer_setup(char *s)
+{
+	pmtmr_ioport = 0;
+	return 1;
+}
+
+__setup("nopmtimer", nopmtimer_setup);

arch/x86/kernel/process_64.c

@@ -0,0 +1,903 @@
+/*
+ *  linux/arch/x86-64/kernel/process.c
+ *
+ *  Copyright (C) 1995  Linus Torvalds
+ *
+ *  Pentium III FXSR, SSE support
+ *	Gareth Hughes <gareth@valinux.com>, May 2000
+ * 
+ *  X86-64 port
+ *	Andi Kleen.
+ *
+ *	CPU hotplug support - ashok.raj@intel.com
+ */
+
+/*
+ * This file handles the architecture-dependent parts of process handling..
+ */
+
+#include <stdarg.h>
+
+#include <linux/cpu.h>
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/fs.h>
+#include <linux/elfcore.h>
+#include <linux/smp.h>
+#include <linux/slab.h>
+#include <linux/user.h>
+#include <linux/module.h>
+#include <linux/a.out.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/ptrace.h>
+#include <linux/utsname.h>
+#include <linux/random.h>
+#include <linux/notifier.h>
+#include <linux/kprobes.h>
+#include <linux/kdebug.h>
+
+#include <asm/uaccess.h>
+#include <asm/pgtable.h>
+#include <asm/system.h>
+#include <asm/io.h>
+#include <asm/processor.h>
+#include <asm/i387.h>
+#include <asm/mmu_context.h>
+#include <asm/pda.h>
+#include <asm/prctl.h>
+#include <asm/desc.h>
+#include <asm/proto.h>
+#include <asm/ia32.h>
+#include <asm/idle.h>
+
+asmlinkage extern void ret_from_fork(void);
+
+unsigned long kernel_thread_flags = CLONE_VM | CLONE_UNTRACED;
+
+unsigned long boot_option_idle_override = 0;
+EXPORT_SYMBOL(boot_option_idle_override);
+
+/*
+ * Powermanagement idle function, if any..
+ */
+void (*pm_idle)(void);
+EXPORT_SYMBOL(pm_idle);
+static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
+
+static ATOMIC_NOTIFIER_HEAD(idle_notifier);
+
+void idle_notifier_register(struct notifier_block *n)
+{
+	atomic_notifier_chain_register(&idle_notifier, n);
+}
+EXPORT_SYMBOL_GPL(idle_notifier_register);
+
+void idle_notifier_unregister(struct notifier_block *n)
+{
+	atomic_notifier_chain_unregister(&idle_notifier, n);
+}
+EXPORT_SYMBOL(idle_notifier_unregister);
+
+void enter_idle(void)
+{
+	write_pda(isidle, 1);
+	atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL);
+}
+
+static void __exit_idle(void)
+{
+	if (test_and_clear_bit_pda(0, isidle) == 0)
+		return;
+	atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL);
+}
+
+/* Called from interrupts to signify idle end */
+void exit_idle(void)
+{
+	/* idle loop has pid 0 */
+	if (current->pid)
+		return;
+	__exit_idle();
+}
+
+/*
+ * We use this if we don't have any better
+ * idle routine..
+ */
+static void default_idle(void)
+{
+	current_thread_info()->status &= ~TS_POLLING;
+	/*
+	 * TS_POLLING-cleared state must be visible before we
+	 * test NEED_RESCHED:
+	 */
+	smp_mb();
+	local_irq_disable();
+	if (!need_resched()) {
+		/* Enables interrupts one instruction before HLT.
+		   x86 special cases this so there is no race. */
+		safe_halt();
+	} else
+		local_irq_enable();
+	current_thread_info()->status |= TS_POLLING;
+}
+
+/*
+ * On SMP it's slightly faster (but much more power-consuming!)
+ * to poll the ->need_resched flag instead of waiting for the
+ * cross-CPU IPI to arrive. Use this option with caution.
+ */
+static void poll_idle (void)
+{
+	local_irq_enable();
+	cpu_relax();
+}
+
+void cpu_idle_wait(void)
+{
+	unsigned int cpu, this_cpu = get_cpu();
+	cpumask_t map, tmp = current->cpus_allowed;
+
+	set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
+	put_cpu();
+
+	cpus_clear(map);
+	for_each_online_cpu(cpu) {
+		per_cpu(cpu_idle_state, cpu) = 1;
+		cpu_set(cpu, map);
+	}
+
+	__get_cpu_var(cpu_idle_state) = 0;
+
+	wmb();
+	do {
+		ssleep(1);
+		for_each_online_cpu(cpu) {
+			if (cpu_isset(cpu, map) &&
+					!per_cpu(cpu_idle_state, cpu))
+				cpu_clear(cpu, map);
+		}
+		cpus_and(map, map, cpu_online_map);
+	} while (!cpus_empty(map));
+
+	set_cpus_allowed(current, tmp);
+}
+EXPORT_SYMBOL_GPL(cpu_idle_wait);
+
+#ifdef CONFIG_HOTPLUG_CPU
+DECLARE_PER_CPU(int, cpu_state);
+
+#include <asm/nmi.h>
+/* We halt the CPU with physical CPU hotplug */
+static inline void play_dead(void)
+{
+	idle_task_exit();
+	wbinvd();
+	mb();
+	/* Ack it */
+	__get_cpu_var(cpu_state) = CPU_DEAD;
+
+	local_irq_disable();
+	while (1)
+		halt();
+}
+#else
+static inline void play_dead(void)
+{
+	BUG();
+}
+#endif /* CONFIG_HOTPLUG_CPU */
+
+/*
+ * The idle thread. There's no useful work to be
+ * done, so just try to conserve power and have a
+ * low exit latency (ie sit in a loop waiting for
+ * somebody to say that they'd like to reschedule)
+ */
+void cpu_idle (void)
+{
+	current_thread_info()->status |= TS_POLLING;
+	/* endless idle loop with no priority at all */
+	while (1) {
+		while (!need_resched()) {
+			void (*idle)(void);
+
+			if (__get_cpu_var(cpu_idle_state))
+				__get_cpu_var(cpu_idle_state) = 0;
+
+			rmb();
+			idle = pm_idle;
+			if (!idle)
+				idle = default_idle;
+			if (cpu_is_offline(smp_processor_id()))
+				play_dead();
+			/*
+			 * Idle routines should keep interrupts disabled
+			 * from here on, until they go to idle.
+			 * Otherwise, idle callbacks can misfire.
+			 */
+			local_irq_disable();
+			enter_idle();
+			idle();
+			/* In many cases the interrupt that ended idle
+			   has already called exit_idle. But some idle
+			   loops can be woken up without interrupt. */
+			__exit_idle();
+		}
+
+		preempt_enable_no_resched();
+		schedule();
+		preempt_disable();
+	}
+}
+
+/*
+ * This uses new MONITOR/MWAIT instructions on P4 processors with PNI,
+ * which can obviate IPI to trigger checking of need_resched.
+ * We execute MONITOR against need_resched and enter optimized wait state
+ * through MWAIT. Whenever someone changes need_resched, we would be woken
+ * up from MWAIT (without an IPI).
+ *
+ * New with Core Duo processors, MWAIT can take some hints based on CPU
+ * capability.
+ */
+void mwait_idle_with_hints(unsigned long eax, unsigned long ecx)
+{
+	if (!need_resched()) {
+		__monitor((void *)&current_thread_info()->flags, 0, 0);
+		smp_mb();
+		if (!need_resched())
+			__mwait(eax, ecx);
+	}
+}
+
+/* Default MONITOR/MWAIT with no hints, used for default C1 state */
+static void mwait_idle(void)
+{
+	if (!need_resched()) {
+		__monitor((void *)&current_thread_info()->flags, 0, 0);
+		smp_mb();
+		if (!need_resched())
+			__sti_mwait(0, 0);
+		else
+			local_irq_enable();
+	} else {
+		local_irq_enable();
+	}
+}
+
+void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
+{
+	static int printed;
+	if (cpu_has(c, X86_FEATURE_MWAIT)) {
+		/*
+		 * Skip, if setup has overridden idle.
+		 * One CPU supports mwait => All CPUs supports mwait
+		 */
+		if (!pm_idle) {
+			if (!printed) {
+				printk(KERN_INFO "using mwait in idle threads.\n");
+				printed = 1;
+			}
+			pm_idle = mwait_idle;
+		}
+	}
+}
+
+static int __init idle_setup (char *str)
+{
+	if (!strcmp(str, "poll")) {
+		printk("using polling idle threads.\n");
+		pm_idle = poll_idle;
+	} else if (!strcmp(str, "mwait"))
+		force_mwait = 1;
+	else
+		return -1;
+
+	boot_option_idle_override = 1;
+	return 0;
+}
+early_param("idle", idle_setup);
+
+/* Prints also some state that isn't saved in the pt_regs */ 
+void __show_regs(struct pt_regs * regs)
+{
+	unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L, fs, gs, shadowgs;
+	unsigned long d0, d1, d2, d3, d6, d7;
+	unsigned int fsindex,gsindex;
+	unsigned int ds,cs,es; 
+
+	printk("\n");
+	print_modules();
+	printk("Pid: %d, comm: %.20s %s %s %.*s\n",
+		current->pid, current->comm, print_tainted(),
+		init_utsname()->release,
+		(int)strcspn(init_utsname()->version, " "),
+		init_utsname()->version);
+	printk("RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->rip);
+	printk_address(regs->rip); 
+	printk("RSP: %04lx:%016lx  EFLAGS: %08lx\n", regs->ss, regs->rsp,
+		regs->eflags);
+	printk("RAX: %016lx RBX: %016lx RCX: %016lx\n",
+	       regs->rax, regs->rbx, regs->rcx);
+	printk("RDX: %016lx RSI: %016lx RDI: %016lx\n",
+	       regs->rdx, regs->rsi, regs->rdi); 
+	printk("RBP: %016lx R08: %016lx R09: %016lx\n",
+	       regs->rbp, regs->r8, regs->r9); 
+	printk("R10: %016lx R11: %016lx R12: %016lx\n",
+	       regs->r10, regs->r11, regs->r12); 
+	printk("R13: %016lx R14: %016lx R15: %016lx\n",
+	       regs->r13, regs->r14, regs->r15); 
+
+	asm("movl %%ds,%0" : "=r" (ds)); 
+	asm("movl %%cs,%0" : "=r" (cs)); 
+	asm("movl %%es,%0" : "=r" (es)); 
+	asm("movl %%fs,%0" : "=r" (fsindex));
+	asm("movl %%gs,%0" : "=r" (gsindex));
+
+	rdmsrl(MSR_FS_BASE, fs);
+	rdmsrl(MSR_GS_BASE, gs); 
+	rdmsrl(MSR_KERNEL_GS_BASE, shadowgs); 
+
+	cr0 = read_cr0();
+	cr2 = read_cr2();
+	cr3 = read_cr3();
+	cr4 = read_cr4();
+
+	printk("FS:  %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n", 
+	       fs,fsindex,gs,gsindex,shadowgs); 
+	printk("CS:  %04x DS: %04x ES: %04x CR0: %016lx\n", cs, ds, es, cr0); 
+	printk("CR2: %016lx CR3: %016lx CR4: %016lx\n", cr2, cr3, cr4);
+
+	get_debugreg(d0, 0);
+	get_debugreg(d1, 1);
+	get_debugreg(d2, 2);
+	printk("DR0: %016lx DR1: %016lx DR2: %016lx\n", d0, d1, d2);
+	get_debugreg(d3, 3);
+	get_debugreg(d6, 6);
+	get_debugreg(d7, 7);
+	printk("DR3: %016lx DR6: %016lx DR7: %016lx\n", d3, d6, d7);
+}
+
+void show_regs(struct pt_regs *regs)
+{
+	printk("CPU %d:", smp_processor_id());
+	__show_regs(regs);
+	show_trace(NULL, regs, (void *)(regs + 1));
+}
+
+/*
+ * Free current thread data structures etc..
+ */
+void exit_thread(void)
+{
+	struct task_struct *me = current;
+	struct thread_struct *t = &me->thread;
+
+	if (me->thread.io_bitmap_ptr) { 
+		struct tss_struct *tss = &per_cpu(init_tss, get_cpu());
+
+		kfree(t->io_bitmap_ptr);
+		t->io_bitmap_ptr = NULL;
+		clear_thread_flag(TIF_IO_BITMAP);
+		/*
+		 * Careful, clear this in the TSS too:
+		 */
+		memset(tss->io_bitmap, 0xff, t->io_bitmap_max);
+		t->io_bitmap_max = 0;
+		put_cpu();
+	}
+}
+
+void flush_thread(void)
+{
+	struct task_struct *tsk = current;
+
+	if (test_tsk_thread_flag(tsk, TIF_ABI_PENDING)) {
+		clear_tsk_thread_flag(tsk, TIF_ABI_PENDING);
+		if (test_tsk_thread_flag(tsk, TIF_IA32)) {
+			clear_tsk_thread_flag(tsk, TIF_IA32);
+		} else {
+			set_tsk_thread_flag(tsk, TIF_IA32);
+			current_thread_info()->status |= TS_COMPAT;
+		}
+	}
+	clear_tsk_thread_flag(tsk, TIF_DEBUG);
+
+	tsk->thread.debugreg0 = 0;
+	tsk->thread.debugreg1 = 0;
+	tsk->thread.debugreg2 = 0;
+	tsk->thread.debugreg3 = 0;
+	tsk->thread.debugreg6 = 0;
+	tsk->thread.debugreg7 = 0;
+	memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array));	
+	/*
+	 * Forget coprocessor state..
+	 */
+	clear_fpu(tsk);
+	clear_used_math();
+}
+
+void release_thread(struct task_struct *dead_task)
+{
+	if (dead_task->mm) {
+		if (dead_task->mm->context.size) {
+			printk("WARNING: dead process %8s still has LDT? <%p/%d>\n",
+					dead_task->comm,
+					dead_task->mm->context.ldt,
+					dead_task->mm->context.size);
+			BUG();
+		}
+	}
+}
+
+static inline void set_32bit_tls(struct task_struct *t, int tls, u32 addr)
+{
+	struct user_desc ud = { 
+		.base_addr = addr,
+		.limit = 0xfffff,
+		.seg_32bit = 1,
+		.limit_in_pages = 1,
+		.useable = 1,
+	};
+	struct n_desc_struct *desc = (void *)t->thread.tls_array;
+	desc += tls;
+	desc->a = LDT_entry_a(&ud); 
+	desc->b = LDT_entry_b(&ud); 
+}
+
+static inline u32 read_32bit_tls(struct task_struct *t, int tls)
+{
+	struct desc_struct *desc = (void *)t->thread.tls_array;
+	desc += tls;
+	return desc->base0 | 
+		(((u32)desc->base1) << 16) | 
+		(((u32)desc->base2) << 24);
+}
+
+/*
+ * This gets called before we allocate a new thread and copy
+ * the current task into it.
+ */
+void prepare_to_copy(struct task_struct *tsk)
+{
+	unlazy_fpu(tsk);
+}
+
+int copy_thread(int nr, unsigned long clone_flags, unsigned long rsp, 
+		unsigned long unused,
+	struct task_struct * p, struct pt_regs * regs)
+{
+	int err;
+	struct pt_regs * childregs;
+	struct task_struct *me = current;
+
+	childregs = ((struct pt_regs *)
+			(THREAD_SIZE + task_stack_page(p))) - 1;
+	*childregs = *regs;
+
+	childregs->rax = 0;
+	childregs->rsp = rsp;
+	if (rsp == ~0UL)
+		childregs->rsp = (unsigned long)childregs;
+
+	p->thread.rsp = (unsigned long) childregs;
+	p->thread.rsp0 = (unsigned long) (childregs+1);
+	p->thread.userrsp = me->thread.userrsp; 
+
+	set_tsk_thread_flag(p, TIF_FORK);
+
+	p->thread.fs = me->thread.fs;
+	p->thread.gs = me->thread.gs;
+
+	asm("mov %%gs,%0" : "=m" (p->thread.gsindex));
+	asm("mov %%fs,%0" : "=m" (p->thread.fsindex));
+	asm("mov %%es,%0" : "=m" (p->thread.es));
+	asm("mov %%ds,%0" : "=m" (p->thread.ds));
+
+	if (unlikely(test_tsk_thread_flag(me, TIF_IO_BITMAP))) {
+		p->thread.io_bitmap_ptr = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
+		if (!p->thread.io_bitmap_ptr) {
+			p->thread.io_bitmap_max = 0;
+			return -ENOMEM;
+		}
+		memcpy(p->thread.io_bitmap_ptr, me->thread.io_bitmap_ptr,
+				IO_BITMAP_BYTES);
+		set_tsk_thread_flag(p, TIF_IO_BITMAP);
+	} 
+
+	/*
+	 * Set a new TLS for the child thread?
+	 */
+	if (clone_flags & CLONE_SETTLS) {
+#ifdef CONFIG_IA32_EMULATION
+		if (test_thread_flag(TIF_IA32))
+			err = ia32_child_tls(p, childregs); 
+		else 			
+#endif	 
+			err = do_arch_prctl(p, ARCH_SET_FS, childregs->r8); 
+		if (err) 
+			goto out;
+	}
+	err = 0;
+out:
+	if (err && p->thread.io_bitmap_ptr) {
+		kfree(p->thread.io_bitmap_ptr);
+		p->thread.io_bitmap_max = 0;
+	}
+	return err;
+}
+
+/*
+ * This special macro can be used to load a debugging register
+ */
+#define loaddebug(thread,r) set_debugreg(thread->debugreg ## r, r)
+
+static inline void __switch_to_xtra(struct task_struct *prev_p,
+			     	    struct task_struct *next_p,
+			     	    struct tss_struct *tss)
+{
+	struct thread_struct *prev, *next;
+
+	prev = &prev_p->thread,
+	next = &next_p->thread;
+
+	if (test_tsk_thread_flag(next_p, TIF_DEBUG)) {
+		loaddebug(next, 0);
+		loaddebug(next, 1);
+		loaddebug(next, 2);
+		loaddebug(next, 3);
+		/* no 4 and 5 */
+		loaddebug(next, 6);
+		loaddebug(next, 7);
+	}
+
+	if (test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
+		/*
+		 * Copy the relevant range of the IO bitmap.
+		 * Normally this is 128 bytes or less:
+		 */
+		memcpy(tss->io_bitmap, next->io_bitmap_ptr,
+		       max(prev->io_bitmap_max, next->io_bitmap_max));
+	} else if (test_tsk_thread_flag(prev_p, TIF_IO_BITMAP)) {
+		/*
+		 * Clear any possible leftover bits:
+		 */
+		memset(tss->io_bitmap, 0xff, prev->io_bitmap_max);
+	}
+}
+
+/*
+ *	switch_to(x,y) should switch tasks from x to y.
+ *
+ * This could still be optimized: 
+ * - fold all the options into a flag word and test it with a single test.
+ * - could test fs/gs bitsliced
+ *
+ * Kprobes not supported here. Set the probe on schedule instead.
+ */
+__kprobes struct task_struct *
+__switch_to(struct task_struct *prev_p, struct task_struct *next_p)
+{
+	struct thread_struct *prev = &prev_p->thread,
+				 *next = &next_p->thread;
+	int cpu = smp_processor_id();  
+	struct tss_struct *tss = &per_cpu(init_tss, cpu);
+
+	/* we're going to use this soon, after a few expensive things */
+	if (next_p->fpu_counter>5)
+		prefetch(&next->i387.fxsave);
+
+	/*
+	 * Reload esp0, LDT and the page table pointer:
+	 */
+	tss->rsp0 = next->rsp0;
+
+	/* 
+	 * Switch DS and ES.
+	 * This won't pick up thread selector changes, but I guess that is ok.
+	 */
+	asm volatile("mov %%es,%0" : "=m" (prev->es));
+	if (unlikely(next->es | prev->es))
+		loadsegment(es, next->es); 
+	
+	asm volatile ("mov %%ds,%0" : "=m" (prev->ds));
+	if (unlikely(next->ds | prev->ds))
+		loadsegment(ds, next->ds);
+
+	load_TLS(next, cpu);
+
+	/* 
+	 * Switch FS and GS.
+	 */
+	{ 
+		unsigned fsindex;
+		asm volatile("movl %%fs,%0" : "=r" (fsindex)); 
+		/* segment register != 0 always requires a reload. 
+		   also reload when it has changed. 
+		   when prev process used 64bit base always reload
+		   to avoid an information leak. */
+		if (unlikely(fsindex | next->fsindex | prev->fs)) {
+			loadsegment(fs, next->fsindex);
+			/* check if the user used a selector != 0
+	                 * if yes clear 64bit base, since overloaded base
+                         * is always mapped to the Null selector
+                         */
+			if (fsindex)
+			prev->fs = 0;				
+		}
+		/* when next process has a 64bit base use it */
+		if (next->fs) 
+			wrmsrl(MSR_FS_BASE, next->fs); 
+		prev->fsindex = fsindex;
+	}
+	{ 
+		unsigned gsindex;
+		asm volatile("movl %%gs,%0" : "=r" (gsindex)); 
+		if (unlikely(gsindex | next->gsindex | prev->gs)) {
+			load_gs_index(next->gsindex);
+			if (gsindex)
+			prev->gs = 0;				
+		}
+		if (next->gs)
+			wrmsrl(MSR_KERNEL_GS_BASE, next->gs); 
+		prev->gsindex = gsindex;
+	}
+
+	/* Must be after DS reload */
+	unlazy_fpu(prev_p);
+
+	/* 
+	 * Switch the PDA and FPU contexts.
+	 */
+	prev->userrsp = read_pda(oldrsp); 
+	write_pda(oldrsp, next->userrsp); 
+	write_pda(pcurrent, next_p); 
+
+	write_pda(kernelstack,
+	(unsigned long)task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);
+#ifdef CONFIG_CC_STACKPROTECTOR
+	write_pda(stack_canary, next_p->stack_canary);
+	/*
+	 * Build time only check to make sure the stack_canary is at
+	 * offset 40 in the pda; this is a gcc ABI requirement
+	 */
+	BUILD_BUG_ON(offsetof(struct x8664_pda, stack_canary) != 40);
+#endif
+
+	/*
+	 * Now maybe reload the debug registers and handle I/O bitmaps
+	 */
+	if (unlikely((task_thread_info(next_p)->flags & _TIF_WORK_CTXSW))
+	    || test_tsk_thread_flag(prev_p, TIF_IO_BITMAP))
+		__switch_to_xtra(prev_p, next_p, tss);
+
+	/* If the task has used fpu the last 5 timeslices, just do a full
+	 * restore of the math state immediately to avoid the trap; the
+	 * chances of needing FPU soon are obviously high now
+	 */
+	if (next_p->fpu_counter>5)
+		math_state_restore();
+	return prev_p;
+}
+
+/*
+ * sys_execve() executes a new program.
+ */
+asmlinkage 
+long sys_execve(char __user *name, char __user * __user *argv,
+		char __user * __user *envp, struct pt_regs regs)
+{
+	long error;
+	char * filename;
+
+	filename = getname(name);
+	error = PTR_ERR(filename);
+	if (IS_ERR(filename)) 
+		return error;
+	error = do_execve(filename, argv, envp, &regs); 
+	if (error == 0) {
+		task_lock(current);
+		current->ptrace &= ~PT_DTRACE;
+		task_unlock(current);
+	}
+	putname(filename);
+	return error;
+}
+
+void set_personality_64bit(void)
+{
+	/* inherit personality from parent */
+
+	/* Make sure to be in 64bit mode */
+	clear_thread_flag(TIF_IA32); 
+
+	/* TBD: overwrites user setup. Should have two bits.
+	   But 64bit processes have always behaved this way,
+	   so it's not too bad. The main problem is just that
+   	   32bit childs are affected again. */
+	current->personality &= ~READ_IMPLIES_EXEC;
+}
+
+asmlinkage long sys_fork(struct pt_regs *regs)
+{
+	return do_fork(SIGCHLD, regs->rsp, regs, 0, NULL, NULL);
+}
+
+asmlinkage long
+sys_clone(unsigned long clone_flags, unsigned long newsp,
+	  void __user *parent_tid, void __user *child_tid, struct pt_regs *regs)
+{
+	if (!newsp)
+		newsp = regs->rsp;
+	return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid);
+}
+
+/*
+ * This is trivial, and on the face of it looks like it
+ * could equally well be done in user mode.
+ *
+ * Not so, for quite unobvious reasons - register pressure.
+ * In user mode vfork() cannot have a stack frame, and if
+ * done by calling the "clone()" system call directly, you
+ * do not have enough call-clobbered registers to hold all
+ * the information you need.
+ */
+asmlinkage long sys_vfork(struct pt_regs *regs)
+{
+	return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->rsp, regs, 0,
+		    NULL, NULL);
+}
+
+unsigned long get_wchan(struct task_struct *p)
+{
+	unsigned long stack;
+	u64 fp,rip;
+	int count = 0;
+
+	if (!p || p == current || p->state==TASK_RUNNING)
+		return 0; 
+	stack = (unsigned long)task_stack_page(p);
+	if (p->thread.rsp < stack || p->thread.rsp > stack+THREAD_SIZE)
+		return 0;
+	fp = *(u64 *)(p->thread.rsp);
+	do { 
+		if (fp < (unsigned long)stack ||
+		    fp > (unsigned long)stack+THREAD_SIZE)
+			return 0; 
+		rip = *(u64 *)(fp+8); 
+		if (!in_sched_functions(rip))
+			return rip; 
+		fp = *(u64 *)fp; 
+	} while (count++ < 16); 
+	return 0;
+}
+
+long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
+{ 
+	int ret = 0; 
+	int doit = task == current;
+	int cpu;
+
+	switch (code) { 
+	case ARCH_SET_GS:
+		if (addr >= TASK_SIZE_OF(task))
+			return -EPERM; 
+		cpu = get_cpu();
+		/* handle small bases via the GDT because that's faster to 
+		   switch. */
+		if (addr <= 0xffffffff) {  
+			set_32bit_tls(task, GS_TLS, addr); 
+			if (doit) { 
+				load_TLS(&task->thread, cpu);
+				load_gs_index(GS_TLS_SEL); 
+			}
+			task->thread.gsindex = GS_TLS_SEL; 
+			task->thread.gs = 0;
+		} else { 
+			task->thread.gsindex = 0;
+			task->thread.gs = addr;
+			if (doit) {
+				load_gs_index(0);
+				ret = checking_wrmsrl(MSR_KERNEL_GS_BASE, addr);
+			} 
+		}
+		put_cpu();
+		break;
+	case ARCH_SET_FS:
+		/* Not strictly needed for fs, but do it for symmetry
+		   with gs */
+		if (addr >= TASK_SIZE_OF(task))
+			return -EPERM; 
+		cpu = get_cpu();
+		/* handle small bases via the GDT because that's faster to 
+		   switch. */
+		if (addr <= 0xffffffff) { 
+			set_32bit_tls(task, FS_TLS, addr);
+			if (doit) { 
+				load_TLS(&task->thread, cpu); 
+				asm volatile("movl %0,%%fs" :: "r"(FS_TLS_SEL));
+			}
+			task->thread.fsindex = FS_TLS_SEL;
+			task->thread.fs = 0;
+		} else { 
+			task->thread.fsindex = 0;
+			task->thread.fs = addr;
+			if (doit) {
+				/* set the selector to 0 to not confuse
+				   __switch_to */
+				asm volatile("movl %0,%%fs" :: "r" (0));
+				ret = checking_wrmsrl(MSR_FS_BASE, addr);
+			}
+		}
+		put_cpu();
+		break;
+	case ARCH_GET_FS: { 
+		unsigned long base; 
+		if (task->thread.fsindex == FS_TLS_SEL)
+			base = read_32bit_tls(task, FS_TLS);
+		else if (doit)
+			rdmsrl(MSR_FS_BASE, base);
+		else
+			base = task->thread.fs;
+		ret = put_user(base, (unsigned long __user *)addr); 
+		break; 
+	}
+	case ARCH_GET_GS: { 
+		unsigned long base;
+		unsigned gsindex;
+		if (task->thread.gsindex == GS_TLS_SEL)
+			base = read_32bit_tls(task, GS_TLS);
+		else if (doit) {
+ 			asm("movl %%gs,%0" : "=r" (gsindex));
+			if (gsindex)
+				rdmsrl(MSR_KERNEL_GS_BASE, base);
+			else
+				base = task->thread.gs;
+		}
+		else
+			base = task->thread.gs;
+		ret = put_user(base, (unsigned long __user *)addr); 
+		break;
+	}
+
+	default:
+		ret = -EINVAL;
+		break;
+	} 
+
+	return ret;	
+} 
+
+long sys_arch_prctl(int code, unsigned long addr)
+{
+	return do_arch_prctl(current, code, addr);
+} 
+
+/* 
+ * Capture the user space registers if the task is not running (in user space)
+ */
+int dump_task_regs(struct task_struct *tsk, elf_gregset_t *regs)
+{
+	struct pt_regs *pp, ptregs;
+
+	pp = task_pt_regs(tsk);
+
+	ptregs = *pp; 
+	ptregs.cs &= 0xffff;
+	ptregs.ss &= 0xffff;
+
+	elf_core_copy_regs(regs, &ptregs);
+ 
+	return 1;
+}
+
+unsigned long arch_align_stack(unsigned long sp)
+{
+	if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
+		sp -= get_random_int() % 8192;
+	return sp & ~0xf;
+}

arch/x86/kernel/ptrace_64.c

@@ -0,0 +1,627 @@
+/* ptrace.c */
+/* By Ross Biro 1/23/92 */
+/*
+ * Pentium III FXSR, SSE support
+ *	Gareth Hughes <gareth@valinux.com>, May 2000
+ * 
+ * x86-64 port 2000-2002 Andi Kleen
+ */
+
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/errno.h>
+#include <linux/ptrace.h>
+#include <linux/user.h>
+#include <linux/security.h>
+#include <linux/audit.h>
+#include <linux/seccomp.h>
+#include <linux/signal.h>
+
+#include <asm/uaccess.h>
+#include <asm/pgtable.h>
+#include <asm/system.h>
+#include <asm/processor.h>
+#include <asm/i387.h>
+#include <asm/debugreg.h>
+#include <asm/ldt.h>
+#include <asm/desc.h>
+#include <asm/proto.h>
+#include <asm/ia32.h>
+
+/*
+ * does not yet catch signals sent when the child dies.
+ * in exit.c or in signal.c.
+ */
+
+/*
+ * Determines which flags the user has access to [1 = access, 0 = no access].
+ * Prohibits changing ID(21), VIP(20), VIF(19), VM(17), IOPL(12-13), IF(9).
+ * Also masks reserved bits (63-22, 15, 5, 3, 1).
+ */
+#define FLAG_MASK 0x54dd5UL
+
+/* set's the trap flag. */
+#define TRAP_FLAG 0x100UL
+
+/*
+ * eflags and offset of eflags on child stack..
+ */
+#define EFLAGS offsetof(struct pt_regs, eflags)
+#define EFL_OFFSET ((int)(EFLAGS-sizeof(struct pt_regs)))
+
+/*
+ * this routine will get a word off of the processes privileged stack. 
+ * the offset is how far from the base addr as stored in the TSS.  
+ * this routine assumes that all the privileged stacks are in our
+ * data space.
+ */   
+static inline unsigned long get_stack_long(struct task_struct *task, int offset)
+{
+	unsigned char *stack;
+
+	stack = (unsigned char *)task->thread.rsp0;
+	stack += offset;
+	return (*((unsigned long *)stack));
+}
+
+/*
+ * this routine will put a word on the processes privileged stack. 
+ * the offset is how far from the base addr as stored in the TSS.  
+ * this routine assumes that all the privileged stacks are in our
+ * data space.
+ */
+static inline long put_stack_long(struct task_struct *task, int offset,
+	unsigned long data)
+{
+	unsigned char * stack;
+
+	stack = (unsigned char *) task->thread.rsp0;
+	stack += offset;
+	*(unsigned long *) stack = data;
+	return 0;
+}
+
+#define LDT_SEGMENT 4
+
+unsigned long convert_rip_to_linear(struct task_struct *child, struct pt_regs *regs)
+{
+	unsigned long addr, seg;
+
+	addr = regs->rip;
+	seg = regs->cs & 0xffff;
+
+	/*
+	 * We'll assume that the code segments in the GDT
+	 * are all zero-based. That is largely true: the
+	 * TLS segments are used for data, and the PNPBIOS
+	 * and APM bios ones we just ignore here.
+	 */
+	if (seg & LDT_SEGMENT) {
+		u32 *desc;
+		unsigned long base;
+
+		seg &= ~7UL;
+
+		down(&child->mm->context.sem);
+		if (unlikely((seg >> 3) >= child->mm->context.size))
+			addr = -1L; /* bogus selector, access would fault */
+		else {
+			desc = child->mm->context.ldt + seg;
+			base = ((desc[0] >> 16) |
+				((desc[1] & 0xff) << 16) |
+				(desc[1] & 0xff000000));
+
+			/* 16-bit code segment? */
+			if (!((desc[1] >> 22) & 1))
+				addr &= 0xffff;
+			addr += base;
+		}
+		up(&child->mm->context.sem);
+	}
+
+	return addr;
+}
+
+static int is_setting_trap_flag(struct task_struct *child, struct pt_regs *regs)
+{
+	int i, copied;
+	unsigned char opcode[15];
+	unsigned long addr = convert_rip_to_linear(child, regs);
+
+	copied = access_process_vm(child, addr, opcode, sizeof(opcode), 0);
+	for (i = 0; i < copied; i++) {
+		switch (opcode[i]) {
+		/* popf and iret */
+		case 0x9d: case 0xcf:
+			return 1;
+
+			/* CHECKME: 64 65 */
+
+		/* opcode and address size prefixes */
+		case 0x66: case 0x67:
+			continue;
+		/* irrelevant prefixes (segment overrides and repeats) */
+		case 0x26: case 0x2e:
+		case 0x36: case 0x3e:
+		case 0x64: case 0x65:
+		case 0xf2: case 0xf3:
+			continue;
+
+		case 0x40 ... 0x4f:
+			if (regs->cs != __USER_CS)
+				/* 32-bit mode: register increment */
+				return 0;
+			/* 64-bit mode: REX prefix */
+			continue;
+
+			/* CHECKME: f2, f3 */
+
+		/*
+		 * pushf: NOTE! We should probably not let
+		 * the user see the TF bit being set. But
+		 * it's more pain than it's worth to avoid
+		 * it, and a debugger could emulate this
+		 * all in user space if it _really_ cares.
+		 */
+		case 0x9c:
+		default:
+			return 0;
+		}
+	}
+	return 0;
+}
+
+static void set_singlestep(struct task_struct *child)
+{
+	struct pt_regs *regs = task_pt_regs(child);
+
+	/*
+	 * Always set TIF_SINGLESTEP - this guarantees that
+	 * we single-step system calls etc..  This will also
+	 * cause us to set TF when returning to user mode.
+	 */
+	set_tsk_thread_flag(child, TIF_SINGLESTEP);
+
+	/*
+	 * If TF was already set, don't do anything else
+	 */
+	if (regs->eflags & TRAP_FLAG)
+		return;
+
+	/* Set TF on the kernel stack.. */
+	regs->eflags |= TRAP_FLAG;
+
+	/*
+	 * ..but if TF is changed by the instruction we will trace,
+	 * don't mark it as being "us" that set it, so that we
+	 * won't clear it by hand later.
+	 */
+	if (is_setting_trap_flag(child, regs))
+		return;
+
+	child->ptrace |= PT_DTRACE;
+}
+
+static void clear_singlestep(struct task_struct *child)
+{
+	/* Always clear TIF_SINGLESTEP... */
+	clear_tsk_thread_flag(child, TIF_SINGLESTEP);
+
+	/* But touch TF only if it was set by us.. */
+	if (child->ptrace & PT_DTRACE) {
+		struct pt_regs *regs = task_pt_regs(child);
+		regs->eflags &= ~TRAP_FLAG;
+		child->ptrace &= ~PT_DTRACE;
+	}
+}
+
+/*
+ * Called by kernel/ptrace.c when detaching..
+ *
+ * Make sure the single step bit is not set.
+ */
+void ptrace_disable(struct task_struct *child)
+{ 
+	clear_singlestep(child);
+}
+
+static int putreg(struct task_struct *child,
+	unsigned long regno, unsigned long value)
+{
+	unsigned long tmp; 
+	
+	switch (regno) {
+		case offsetof(struct user_regs_struct,fs):
+			if (value && (value & 3) != 3)
+				return -EIO;
+			child->thread.fsindex = value & 0xffff; 
+			return 0;
+		case offsetof(struct user_regs_struct,gs):
+			if (value && (value & 3) != 3)
+				return -EIO;
+			child->thread.gsindex = value & 0xffff;
+			return 0;
+		case offsetof(struct user_regs_struct,ds):
+			if (value && (value & 3) != 3)
+				return -EIO;
+			child->thread.ds = value & 0xffff;
+			return 0;
+		case offsetof(struct user_regs_struct,es): 
+			if (value && (value & 3) != 3)
+				return -EIO;
+			child->thread.es = value & 0xffff;
+			return 0;
+		case offsetof(struct user_regs_struct,ss):
+			if ((value & 3) != 3)
+				return -EIO;
+			value &= 0xffff;
+			return 0;
+		case offsetof(struct user_regs_struct,fs_base):
+			if (value >= TASK_SIZE_OF(child))
+				return -EIO;
+			child->thread.fs = value;
+			return 0;
+		case offsetof(struct user_regs_struct,gs_base):
+			if (value >= TASK_SIZE_OF(child))
+				return -EIO;
+			child->thread.gs = value;
+			return 0;
+		case offsetof(struct user_regs_struct, eflags):
+			value &= FLAG_MASK;
+			tmp = get_stack_long(child, EFL_OFFSET); 
+			tmp &= ~FLAG_MASK; 
+			value |= tmp;
+			break;
+		case offsetof(struct user_regs_struct,cs): 
+			if ((value & 3) != 3)
+				return -EIO;
+			value &= 0xffff;
+			break;
+	}
+	put_stack_long(child, regno - sizeof(struct pt_regs), value);
+	return 0;
+}
+
+static unsigned long getreg(struct task_struct *child, unsigned long regno)
+{
+	unsigned long val;
+	switch (regno) {
+		case offsetof(struct user_regs_struct, fs):
+			return child->thread.fsindex;
+		case offsetof(struct user_regs_struct, gs):
+			return child->thread.gsindex;
+		case offsetof(struct user_regs_struct, ds):
+			return child->thread.ds;
+		case offsetof(struct user_regs_struct, es):
+			return child->thread.es; 
+		case offsetof(struct user_regs_struct, fs_base):
+			return child->thread.fs;
+		case offsetof(struct user_regs_struct, gs_base):
+			return child->thread.gs;
+		default:
+			regno = regno - sizeof(struct pt_regs);
+			val = get_stack_long(child, regno);
+			if (test_tsk_thread_flag(child, TIF_IA32))
+				val &= 0xffffffff;
+			return val;
+	}
+
+}
+
+long arch_ptrace(struct task_struct *child, long request, long addr, long data)
+{
+	long i, ret;
+	unsigned ui;
+
+	switch (request) {
+	/* when I and D space are separate, these will need to be fixed. */
+	case PTRACE_PEEKTEXT: /* read word at location addr. */ 
+	case PTRACE_PEEKDATA:
+		ret = generic_ptrace_peekdata(child, addr, data);
+		break;
+
+	/* read the word at location addr in the USER area. */
+	case PTRACE_PEEKUSR: {
+		unsigned long tmp;
+
+		ret = -EIO;
+		if ((addr & 7) ||
+		    addr > sizeof(struct user) - 7)
+			break;
+
+		switch (addr) { 
+		case 0 ... sizeof(struct user_regs_struct) - sizeof(long):
+			tmp = getreg(child, addr);
+			break;
+		case offsetof(struct user, u_debugreg[0]):
+			tmp = child->thread.debugreg0;
+			break;
+		case offsetof(struct user, u_debugreg[1]):
+			tmp = child->thread.debugreg1;
+			break;
+		case offsetof(struct user, u_debugreg[2]):
+			tmp = child->thread.debugreg2;
+			break;
+		case offsetof(struct user, u_debugreg[3]):
+			tmp = child->thread.debugreg3;
+			break;
+		case offsetof(struct user, u_debugreg[6]):
+			tmp = child->thread.debugreg6;
+			break;
+		case offsetof(struct user, u_debugreg[7]):
+			tmp = child->thread.debugreg7;
+			break;
+		default:
+			tmp = 0;
+			break;
+		}
+		ret = put_user(tmp,(unsigned long __user *) data);
+		break;
+	}
+
+	/* when I and D space are separate, this will have to be fixed. */
+	case PTRACE_POKETEXT: /* write the word at location addr. */
+	case PTRACE_POKEDATA:
+		ret = generic_ptrace_pokedata(child, addr, data);
+		break;
+
+	case PTRACE_POKEUSR: /* write the word at location addr in the USER area */
+	{
+		int dsize = test_tsk_thread_flag(child, TIF_IA32) ? 3 : 7;
+		ret = -EIO;
+		if ((addr & 7) ||
+		    addr > sizeof(struct user) - 7)
+			break;
+
+		switch (addr) { 
+		case 0 ... sizeof(struct user_regs_struct) - sizeof(long):
+			ret = putreg(child, addr, data);
+			break;
+		/* Disallows to set a breakpoint into the vsyscall */
+		case offsetof(struct user, u_debugreg[0]):
+			if (data >= TASK_SIZE_OF(child) - dsize) break;
+			child->thread.debugreg0 = data;
+			ret = 0;
+			break;
+		case offsetof(struct user, u_debugreg[1]):
+			if (data >= TASK_SIZE_OF(child) - dsize) break;
+			child->thread.debugreg1 = data;
+			ret = 0;
+			break;
+		case offsetof(struct user, u_debugreg[2]):
+			if (data >= TASK_SIZE_OF(child) - dsize) break;
+			child->thread.debugreg2 = data;
+			ret = 0;
+			break;
+		case offsetof(struct user, u_debugreg[3]):
+			if (data >= TASK_SIZE_OF(child) - dsize) break;
+			child->thread.debugreg3 = data;
+			ret = 0;
+			break;
+		case offsetof(struct user, u_debugreg[6]):
+				  if (data >> 32)
+				break; 
+			child->thread.debugreg6 = data;
+			ret = 0;
+			break;
+		case offsetof(struct user, u_debugreg[7]):
+			/* See arch/i386/kernel/ptrace.c for an explanation of
+			 * this awkward check.*/
+			data &= ~DR_CONTROL_RESERVED;
+			for(i=0; i<4; i++)
+				if ((0x5554 >> ((data >> (16 + 4*i)) & 0xf)) & 1)
+					break;
+			if (i == 4) {
+			  child->thread.debugreg7 = data;
+			  if (data)
+			  	set_tsk_thread_flag(child, TIF_DEBUG);
+			  else
+			  	clear_tsk_thread_flag(child, TIF_DEBUG);
+			  ret = 0;
+		  	}
+		  break;
+		}
+		break;
+	}
+	case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */
+	case PTRACE_CONT:    /* restart after signal. */
+
+		ret = -EIO;
+		if (!valid_signal(data))
+			break;
+		if (request == PTRACE_SYSCALL)
+			set_tsk_thread_flag(child,TIF_SYSCALL_TRACE);
+		else
+			clear_tsk_thread_flag(child,TIF_SYSCALL_TRACE);
+		clear_tsk_thread_flag(child, TIF_SINGLESTEP);
+		child->exit_code = data;
+		/* make sure the single step bit is not set. */
+		clear_singlestep(child);
+		wake_up_process(child);
+		ret = 0;
+		break;
+
+#ifdef CONFIG_IA32_EMULATION
+		/* This makes only sense with 32bit programs. Allow a
+		   64bit debugger to fully examine them too. Better
+		   don't use it against 64bit processes, use
+		   PTRACE_ARCH_PRCTL instead. */
+	case PTRACE_SET_THREAD_AREA: {
+		struct user_desc __user *p;
+		int old; 
+		p = (struct user_desc __user *)data;
+		get_user(old,  &p->entry_number); 
+		put_user(addr, &p->entry_number);
+		ret = do_set_thread_area(&child->thread, p);
+		put_user(old,  &p->entry_number); 
+		break;
+	case PTRACE_GET_THREAD_AREA:
+		p = (struct user_desc __user *)data;
+		get_user(old,  &p->entry_number); 
+		put_user(addr, &p->entry_number);
+		ret = do_get_thread_area(&child->thread, p);
+		put_user(old,  &p->entry_number); 
+		break;
+	} 
+#endif
+		/* normal 64bit interface to access TLS data. 
+		   Works just like arch_prctl, except that the arguments
+		   are reversed. */
+	case PTRACE_ARCH_PRCTL: 
+		ret = do_arch_prctl(child, data, addr);
+		break;
+
+/*
+ * make the child exit.  Best I can do is send it a sigkill. 
+ * perhaps it should be put in the status that it wants to 
+ * exit.
+ */
+	case PTRACE_KILL:
+		ret = 0;
+		if (child->exit_state == EXIT_ZOMBIE)	/* already dead */
+			break;
+		clear_tsk_thread_flag(child, TIF_SINGLESTEP);
+		child->exit_code = SIGKILL;
+		/* make sure the single step bit is not set. */
+		clear_singlestep(child);
+		wake_up_process(child);
+		break;
+
+	case PTRACE_SINGLESTEP:    /* set the trap flag. */
+		ret = -EIO;
+		if (!valid_signal(data))
+			break;
+		clear_tsk_thread_flag(child,TIF_SYSCALL_TRACE);
+		set_singlestep(child);
+		child->exit_code = data;
+		/* give it a chance to run. */
+		wake_up_process(child);
+		ret = 0;
+		break;
+
+	case PTRACE_DETACH:
+		/* detach a process that was attached. */
+		ret = ptrace_detach(child, data);
+		break;
+
+	case PTRACE_GETREGS: { /* Get all gp regs from the child. */
+	  	if (!access_ok(VERIFY_WRITE, (unsigned __user *)data,
+			       sizeof(struct user_regs_struct))) {
+			ret = -EIO;
+			break;
+		}
+		ret = 0;
+		for (ui = 0; ui < sizeof(struct user_regs_struct); ui += sizeof(long)) {
+			ret |= __put_user(getreg(child, ui),(unsigned long __user *) data);
+			data += sizeof(long);
+		}
+		break;
+	}
+
+	case PTRACE_SETREGS: { /* Set all gp regs in the child. */
+		unsigned long tmp;
+	  	if (!access_ok(VERIFY_READ, (unsigned __user *)data,
+			       sizeof(struct user_regs_struct))) {
+			ret = -EIO;
+			break;
+		}
+		ret = 0;
+		for (ui = 0; ui < sizeof(struct user_regs_struct); ui += sizeof(long)) {
+			ret = __get_user(tmp, (unsigned long __user *) data);
+			if (ret)
+				break;
+			ret = putreg(child, ui, tmp);
+			if (ret)
+				break;
+			data += sizeof(long);
+		}
+		break;
+	}
+
+	case PTRACE_GETFPREGS: { /* Get the child extended FPU state. */
+		if (!access_ok(VERIFY_WRITE, (unsigned __user *)data,
+			       sizeof(struct user_i387_struct))) {
+			ret = -EIO;
+			break;
+		}
+		ret = get_fpregs((struct user_i387_struct __user *)data, child);
+		break;
+	}
+
+	case PTRACE_SETFPREGS: { /* Set the child extended FPU state. */
+		if (!access_ok(VERIFY_READ, (unsigned __user *)data,
+			       sizeof(struct user_i387_struct))) {
+			ret = -EIO;
+			break;
+		}
+		set_stopped_child_used_math(child);
+		ret = set_fpregs(child, (struct user_i387_struct __user *)data);
+		break;
+	}
+
+	default:
+		ret = ptrace_request(child, request, addr, data);
+		break;
+	}
+	return ret;
+}
+
+static void syscall_trace(struct pt_regs *regs)
+{
+
+#if 0
+	printk("trace %s rip %lx rsp %lx rax %d origrax %d caller %lx tiflags %x ptrace %x\n",
+	       current->comm,
+	       regs->rip, regs->rsp, regs->rax, regs->orig_rax, __builtin_return_address(0),
+	       current_thread_info()->flags, current->ptrace); 
+#endif
+
+	ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD)
+				? 0x80 : 0));
+	/*
+	 * this isn't the same as continuing with a signal, but it will do
+	 * for normal use.  strace only continues with a signal if the
+	 * stopping signal is not SIGTRAP.  -brl
+	 */
+	if (current->exit_code) {
+		send_sig(current->exit_code, current, 1);
+		current->exit_code = 0;
+	}
+}
+
+asmlinkage void syscall_trace_enter(struct pt_regs *regs)
+{
+	/* do the secure computing check first */
+	secure_computing(regs->orig_rax);
+
+	if (test_thread_flag(TIF_SYSCALL_TRACE)
+	    && (current->ptrace & PT_PTRACED))
+		syscall_trace(regs);
+
+	if (unlikely(current->audit_context)) {
+		if (test_thread_flag(TIF_IA32)) {
+			audit_syscall_entry(AUDIT_ARCH_I386,
+					    regs->orig_rax,
+					    regs->rbx, regs->rcx,
+					    regs->rdx, regs->rsi);
+		} else {
+			audit_syscall_entry(AUDIT_ARCH_X86_64,
+					    regs->orig_rax,
+					    regs->rdi, regs->rsi,
+					    regs->rdx, regs->r10);
+		}
+	}
+}
+
+asmlinkage void syscall_trace_leave(struct pt_regs *regs)
+{
+	if (unlikely(current->audit_context))
+		audit_syscall_exit(AUDITSC_RESULT(regs->rax), regs->rax);
+
+	if ((test_thread_flag(TIF_SYSCALL_TRACE)
+	     || test_thread_flag(TIF_SINGLESTEP))
+	    && (current->ptrace & PT_PTRACED))
+		syscall_trace(regs);
+}

arch/x86/kernel/reboot_64.c

@@ -0,0 +1,171 @@
+/* Various gunk just to reboot the machine. */ 
+#include <linux/module.h>
+#include <linux/reboot.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/kernel.h>
+#include <linux/ctype.h>
+#include <linux/string.h>
+#include <linux/pm.h>
+#include <linux/kdebug.h>
+#include <linux/sched.h>
+#include <asm/io.h>
+#include <asm/delay.h>
+#include <asm/hw_irq.h>
+#include <asm/system.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+#include <asm/apic.h>
+#include <asm/iommu.h>
+
+/*
+ * Power off function, if any
+ */
+void (*pm_power_off)(void);
+EXPORT_SYMBOL(pm_power_off);
+
+static long no_idt[3];
+static enum { 
+	BOOT_TRIPLE = 't',
+	BOOT_KBD = 'k'
+} reboot_type = BOOT_KBD;
+static int reboot_mode = 0;
+int reboot_force;
+
+/* reboot=t[riple] | k[bd] [, [w]arm | [c]old]
+   warm   Don't set the cold reboot flag
+   cold   Set the cold reboot flag
+   triple Force a triple fault (init)
+   kbd    Use the keyboard controller. cold reset (default)
+   force  Avoid anything that could hang.
+ */ 
+static int __init reboot_setup(char *str)
+{
+	for (;;) {
+		switch (*str) {
+		case 'w': 
+			reboot_mode = 0x1234;
+			break;
+
+		case 'c':
+			reboot_mode = 0;
+			break;
+
+		case 't':
+		case 'b':
+		case 'k':
+			reboot_type = *str;
+			break;
+		case 'f':
+			reboot_force = 1;
+			break;
+		}
+		if((str = strchr(str,',')) != NULL)
+			str++;
+		else
+			break;
+	}
+	return 1;
+}
+
+__setup("reboot=", reboot_setup);
+
+static inline void kb_wait(void)
+{
+	int i;
+
+	for (i=0; i<0x10000; i++)
+		if ((inb_p(0x64) & 0x02) == 0)
+			break;
+}
+
+void machine_shutdown(void)
+{
+	unsigned long flags;
+
+	/* Stop the cpus and apics */
+#ifdef CONFIG_SMP
+	int reboot_cpu_id;
+
+	/* The boot cpu is always logical cpu 0 */
+	reboot_cpu_id = 0;
+
+	/* Make certain the cpu I'm about to reboot on is online */
+	if (!cpu_isset(reboot_cpu_id, cpu_online_map)) {
+		reboot_cpu_id = smp_processor_id();
+	}
+
+	/* Make certain I only run on the appropriate processor */
+	set_cpus_allowed(current, cpumask_of_cpu(reboot_cpu_id));
+
+	/* O.K Now that I'm on the appropriate processor,
+	 * stop all of the others.
+	 */
+	smp_send_stop();
+#endif
+
+	local_irq_save(flags);
+
+#ifndef CONFIG_SMP
+	disable_local_APIC();
+#endif
+
+	disable_IO_APIC();
+
+	local_irq_restore(flags);
+
+	pci_iommu_shutdown();
+}
+
+void machine_emergency_restart(void)
+{
+	int i;
+
+	/* Tell the BIOS if we want cold or warm reboot */
+	*((unsigned short *)__va(0x472)) = reboot_mode;
+       
+	for (;;) {
+		/* Could also try the reset bit in the Hammer NB */
+		switch (reboot_type) { 
+		case BOOT_KBD:
+		for (i=0; i<10; i++) {
+			kb_wait();
+			udelay(50);
+			outb(0xfe,0x64);         /* pulse reset low */
+			udelay(50);
+		}
+
+		case BOOT_TRIPLE: 
+			__asm__ __volatile__("lidt (%0)": :"r" (&no_idt));
+			__asm__ __volatile__("int3");
+
+			reboot_type = BOOT_KBD;
+			break;
+		}      
+	}      
+}
+
+void machine_restart(char * __unused)
+{
+	printk("machine restart\n");
+
+	if (!reboot_force) {
+		machine_shutdown();
+	}
+	machine_emergency_restart();
+}
+
+void machine_halt(void)
+{
+}
+
+void machine_power_off(void)
+{
+	if (pm_power_off) {
+		if (!reboot_force) {
+			machine_shutdown();
+		}
+		pm_power_off();
+	}
+}
+

arch/x86/kernel/relocate_kernel_64.S

@@ -0,0 +1,276 @@
+/*
+ * relocate_kernel.S - put the kernel image in place to boot
+ * Copyright (C) 2002-2005 Eric Biederman  <ebiederm@xmission.com>
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2.  See the file COPYING for more details.
+ */
+
+#include <linux/linkage.h>
+#include <asm/page.h>
+#include <asm/kexec.h>
+
+/*
+ * Must be relocatable PIC code callable as a C function
+ */
+
+#define PTR(x) (x << 3)
+#define PAGE_ALIGNED (1 << PAGE_SHIFT)
+#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */
+
+	.text
+	.align PAGE_ALIGNED
+	.code64
+	.globl relocate_kernel
+relocate_kernel:
+	/* %rdi indirection_page
+	 * %rsi page_list
+	 * %rdx start address
+	 */
+
+	/* map the control page at its virtual address */
+
+	movq	$0x0000ff8000000000, %r10        /* mask */
+	mov	$(39 - 3), %cl                   /* bits to shift */
+	movq	PTR(VA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PGD)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_PUD_0)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	shrq	$9, %r10
+	sub	$9, %cl
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PUD_0)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_PMD_0)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	shrq	$9, %r10
+	sub	$9, %cl
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PMD_0)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_PTE_0)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	shrq	$9, %r10
+	sub	$9, %cl
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PTE_0)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_CONTROL_PAGE)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	/* identity map the control page at its physical address */
+
+	movq	$0x0000ff8000000000, %r10        /* mask */
+	mov	$(39 - 3), %cl                   /* bits to shift */
+	movq	PTR(PA_CONTROL_PAGE)(%rsi), %r11 /* address to map */
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PGD)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_PUD_1)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	shrq	$9, %r10
+	sub	$9, %cl
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PUD_1)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_PMD_1)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	shrq	$9, %r10
+	sub	$9, %cl
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PMD_1)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_PTE_1)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+	shrq	$9, %r10
+	sub	$9, %cl
+
+	movq	%r11, %r9
+	andq	%r10, %r9
+	shrq	%cl, %r9
+
+	movq	PTR(VA_PTE_1)(%rsi), %r8
+	addq	%r8, %r9
+	movq	PTR(PA_CONTROL_PAGE)(%rsi), %r8
+	orq	$PAGE_ATTR, %r8
+	movq	%r8, (%r9)
+
+relocate_new_kernel:
+	/* %rdi indirection_page
+	 * %rsi page_list
+	 * %rdx start address
+	 */
+
+	/* zero out flags, and disable interrupts */
+	pushq $0
+	popfq
+
+	/* get physical address of control page now */
+	/* this is impossible after page table switch */
+	movq	PTR(PA_CONTROL_PAGE)(%rsi), %r8
+
+	/* get physical address of page table now too */
+	movq	PTR(PA_TABLE_PAGE)(%rsi), %rcx
+
+	/* switch to new set of page tables */
+	movq	PTR(PA_PGD)(%rsi), %r9
+	movq	%r9, %cr3
+
+	/* setup a new stack at the end of the physical control page */
+	lea	4096(%r8), %rsp
+
+	/* jump to identity mapped page */
+	addq	$(identity_mapped - relocate_kernel), %r8
+	pushq	%r8
+	ret
+
+identity_mapped:
+	/* store the start address on the stack */
+	pushq   %rdx
+
+	/* Set cr0 to a known state:
+	 * 31 1 == Paging enabled
+	 * 18 0 == Alignment check disabled
+	 * 16 0 == Write protect disabled
+	 * 3  0 == No task switch
+	 * 2  0 == Don't do FP software emulation.
+	 * 0  1 == Proctected mode enabled
+	 */
+	movq	%cr0, %rax
+	andq	$~((1<<18)|(1<<16)|(1<<3)|(1<<2)), %rax
+	orl	$((1<<31)|(1<<0)), %eax
+	movq	%rax, %cr0
+
+	/* Set cr4 to a known state:
+	 * 10 0 == xmm exceptions disabled
+	 * 9  0 == xmm registers instructions disabled
+	 * 8  0 == performance monitoring counter disabled
+	 * 7  0 == page global disabled
+	 * 6  0 == machine check exceptions disabled
+	 * 5  1 == physical address extension enabled
+	 * 4  0 == page size extensions	disabled
+	 * 3  0 == Debug extensions disabled
+	 * 2  0 == Time stamp disable (disabled)
+	 * 1  0 == Protected mode virtual interrupts disabled
+	 * 0  0 == VME disabled
+	 */
+
+	movq	$((1<<5)), %rax
+	movq	%rax, %cr4
+
+	jmp 1f
+1:
+
+	/* Switch to the identity mapped page tables,
+	 * and flush the TLB.
+	*/
+	movq	%rcx, %cr3
+
+	/* Do the copies */
+	movq	%rdi, %rcx 	/* Put the page_list in %rcx */
+	xorq	%rdi, %rdi
+	xorq	%rsi, %rsi
+	jmp	1f
+
+0:	/* top, read another word for the indirection page */
+
+	movq	(%rbx), %rcx
+	addq	$8,	%rbx
+1:
+	testq	$0x1,	%rcx  /* is it a destination page? */
+	jz	2f
+	movq	%rcx,	%rdi
+	andq	$0xfffffffffffff000, %rdi
+	jmp	0b
+2:
+	testq	$0x2,	%rcx  /* is it an indirection page? */
+	jz	2f
+	movq	%rcx,   %rbx
+	andq	$0xfffffffffffff000, %rbx
+	jmp	0b
+2:
+	testq	$0x4,	%rcx  /* is it the done indicator? */
+	jz	2f
+	jmp	3f
+2:
+	testq	$0x8,	%rcx  /* is it the source indicator? */
+	jz	0b	      /* Ignore it otherwise */
+	movq	%rcx,   %rsi  /* For ever source page do a copy */
+	andq	$0xfffffffffffff000, %rsi
+
+	movq	$512,   %rcx
+	rep ; movsq
+	jmp	0b
+3:
+
+	/* To be certain of avoiding problems with self-modifying code
+	 * I need to execute a serializing instruction here.
+	 * So I flush the TLB by reloading %cr3 here, it's handy,
+	 * and not processor dependent.
+	 */
+	movq	%cr3, %rax
+	movq	%rax, %cr3
+
+	/* set all of the registers to known values */
+	/* leave %rsp alone */
+
+	xorq	%rax, %rax
+	xorq	%rbx, %rbx
+	xorq    %rcx, %rcx
+	xorq    %rdx, %rdx
+	xorq    %rsi, %rsi
+	xorq    %rdi, %rdi
+	xorq    %rbp, %rbp
+	xorq	%r8,  %r8
+	xorq	%r9,  %r9
+	xorq	%r10, %r9
+	xorq	%r11, %r11
+	xorq	%r12, %r12
+	xorq	%r13, %r13
+	xorq	%r14, %r14
+	xorq	%r15, %r15
+
+	ret

arch/x86/kernel/setup64.c

@@ -0,0 +1,289 @@
+/* 
+ * X86-64 specific CPU setup.
+ * Copyright (C) 1995  Linus Torvalds
+ * Copyright 2001, 2002, 2003 SuSE Labs / Andi Kleen.
+ * See setup.c for older changelog.
+ */ 
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/string.h>
+#include <linux/bootmem.h>
+#include <linux/bitops.h>
+#include <linux/module.h>
+#include <asm/bootsetup.h>
+#include <asm/pda.h>
+#include <asm/pgtable.h>
+#include <asm/processor.h>
+#include <asm/desc.h>
+#include <asm/atomic.h>
+#include <asm/mmu_context.h>
+#include <asm/smp.h>
+#include <asm/i387.h>
+#include <asm/percpu.h>
+#include <asm/proto.h>
+#include <asm/sections.h>
+
+char x86_boot_params[BOOT_PARAM_SIZE] __initdata;
+
+cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
+
+struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly;
+EXPORT_SYMBOL(_cpu_pda);
+struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned;
+
+struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
+
+char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned")));
+
+unsigned long __supported_pte_mask __read_mostly = ~0UL;
+static int do_not_nx __cpuinitdata = 0;
+
+/* noexec=on|off
+Control non executable mappings for 64bit processes.
+
+on	Enable(default)
+off	Disable
+*/ 
+static int __init nonx_setup(char *str)
+{
+	if (!str)
+		return -EINVAL;
+	if (!strncmp(str, "on", 2)) {
+                __supported_pte_mask |= _PAGE_NX; 
+ 		do_not_nx = 0; 
+	} else if (!strncmp(str, "off", 3)) {
+		do_not_nx = 1;
+		__supported_pte_mask &= ~_PAGE_NX;
+        }
+	return 0;
+} 
+early_param("noexec", nonx_setup);
+
+int force_personality32 = 0; 
+
+/* noexec32=on|off
+Control non executable heap for 32bit processes.
+To control the stack too use noexec=off
+
+on	PROT_READ does not imply PROT_EXEC for 32bit processes
+off	PROT_READ implies PROT_EXEC (default)
+*/
+static int __init nonx32_setup(char *str)
+{
+	if (!strcmp(str, "on"))
+		force_personality32 &= ~READ_IMPLIES_EXEC;
+	else if (!strcmp(str, "off"))
+		force_personality32 |= READ_IMPLIES_EXEC;
+	return 1;
+}
+__setup("noexec32=", nonx32_setup);
+
+/*
+ * Great future plan:
+ * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data.
+ * Always point %gs to its beginning
+ */
+void __init setup_per_cpu_areas(void)
+{ 
+	int i;
+	unsigned long size;
+
+#ifdef CONFIG_HOTPLUG_CPU
+	prefill_possible_map();
+#endif
+
+	/* Copy section for each CPU (we discard the original) */
+	size = PERCPU_ENOUGH_ROOM;
+
+	printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n", size);
+	for_each_cpu_mask (i, cpu_possible_map) {
+		char *ptr;
+
+		if (!NODE_DATA(cpu_to_node(i))) {
+			printk("cpu with no node %d, num_online_nodes %d\n",
+			       i, num_online_nodes());
+			ptr = alloc_bootmem_pages(size);
+		} else { 
+			ptr = alloc_bootmem_pages_node(NODE_DATA(cpu_to_node(i)), size);
+		}
+		if (!ptr)
+			panic("Cannot allocate cpu data for CPU %d\n", i);
+		cpu_pda(i)->data_offset = ptr - __per_cpu_start;
+		memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
+	}
+} 
+
+void pda_init(int cpu)
+{ 
+	struct x8664_pda *pda = cpu_pda(cpu);
+
+	/* Setup up data that may be needed in __get_free_pages early */
+	asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0)); 
+	/* Memory clobbers used to order PDA accessed */
+	mb();
+	wrmsrl(MSR_GS_BASE, pda);
+	mb();
+
+	pda->cpunumber = cpu; 
+	pda->irqcount = -1;
+	pda->kernelstack = 
+		(unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE; 
+	pda->active_mm = &init_mm;
+	pda->mmu_state = 0;
+
+	if (cpu == 0) {
+		/* others are initialized in smpboot.c */
+		pda->pcurrent = &init_task;
+		pda->irqstackptr = boot_cpu_stack; 
+	} else {
+		pda->irqstackptr = (char *)
+			__get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
+		if (!pda->irqstackptr)
+			panic("cannot allocate irqstack for cpu %d", cpu); 
+	}
+
+
+	pda->irqstackptr += IRQSTACKSIZE-64;
+} 
+
+char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]
+__attribute__((section(".bss.page_aligned")));
+
+extern asmlinkage void ignore_sysret(void);
+
+/* May not be marked __init: used by software suspend */
+void syscall_init(void)
+{
+	/* 
+	 * LSTAR and STAR live in a bit strange symbiosis.
+	 * They both write to the same internal register. STAR allows to set CS/DS
+	 * but only a 32bit target. LSTAR sets the 64bit rip. 	 
+	 */ 
+	wrmsrl(MSR_STAR,  ((u64)__USER32_CS)<<48  | ((u64)__KERNEL_CS)<<32); 
+	wrmsrl(MSR_LSTAR, system_call); 
+	wrmsrl(MSR_CSTAR, ignore_sysret);
+
+#ifdef CONFIG_IA32_EMULATION   		
+	syscall32_cpu_init ();
+#endif
+
+	/* Flags to clear on syscall */
+	wrmsrl(MSR_SYSCALL_MASK, EF_TF|EF_DF|EF_IE|0x3000); 
+}
+
+void __cpuinit check_efer(void)
+{
+	unsigned long efer;
+
+	rdmsrl(MSR_EFER, efer); 
+        if (!(efer & EFER_NX) || do_not_nx) { 
+                __supported_pte_mask &= ~_PAGE_NX; 
+        }       
+}
+
+unsigned long kernel_eflags;
+
+/*
+ * cpu_init() initializes state that is per-CPU. Some data is already
+ * initialized (naturally) in the bootstrap process, such as the GDT
+ * and IDT. We reload them nevertheless, this function acts as a
+ * 'CPU state barrier', nothing should get across.
+ * A lot of state is already set up in PDA init.
+ */
+void __cpuinit cpu_init (void)
+{
+	int cpu = stack_smp_processor_id();
+	struct tss_struct *t = &per_cpu(init_tss, cpu);
+	struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
+	unsigned long v; 
+	char *estacks = NULL; 
+	struct task_struct *me;
+	int i;
+
+	/* CPU 0 is initialised in head64.c */
+	if (cpu != 0) {
+		pda_init(cpu);
+	} else 
+		estacks = boot_exception_stacks; 
+
+	me = current;
+
+	if (cpu_test_and_set(cpu, cpu_initialized))
+		panic("CPU#%d already initialized!\n", cpu);
+
+	printk("Initializing CPU#%d\n", cpu);
+
+	clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
+
+	/*
+	 * Initialize the per-CPU GDT with the boot GDT,
+	 * and set up the GDT descriptor:
+	 */
+	if (cpu)
+ 		memcpy(cpu_gdt(cpu), cpu_gdt_table, GDT_SIZE);
+
+	cpu_gdt_descr[cpu].size = GDT_SIZE;
+	asm volatile("lgdt %0" :: "m" (cpu_gdt_descr[cpu]));
+	asm volatile("lidt %0" :: "m" (idt_descr));
+
+	memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
+	syscall_init();
+
+	wrmsrl(MSR_FS_BASE, 0);
+	wrmsrl(MSR_KERNEL_GS_BASE, 0);
+	barrier(); 
+
+	check_efer();
+
+	/*
+	 * set up and load the per-CPU TSS
+	 */
+	for (v = 0; v < N_EXCEPTION_STACKS; v++) {
+		static const unsigned int order[N_EXCEPTION_STACKS] = {
+			[0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER,
+			[DEBUG_STACK - 1] = DEBUG_STACK_ORDER
+		};
+		if (cpu) {
+			estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]);
+			if (!estacks)
+				panic("Cannot allocate exception stack %ld %d\n",
+				      v, cpu); 
+		}
+		estacks += PAGE_SIZE << order[v];
+		orig_ist->ist[v] = t->ist[v] = (unsigned long)estacks;
+	}
+
+	t->io_bitmap_base = offsetof(struct tss_struct, io_bitmap);
+	/*
+	 * <= is required because the CPU will access up to
+	 * 8 bits beyond the end of the IO permission bitmap.
+	 */
+	for (i = 0; i <= IO_BITMAP_LONGS; i++)
+		t->io_bitmap[i] = ~0UL;
+
+	atomic_inc(&init_mm.mm_count);
+	me->active_mm = &init_mm;
+	if (me->mm)
+		BUG();
+	enter_lazy_tlb(&init_mm, me);
+
+	set_tss_desc(cpu, t);
+	load_TR_desc();
+	load_LDT(&init_mm.context);
+
+	/*
+	 * Clear all 6 debug registers:
+	 */
+
+	set_debugreg(0UL, 0);
+	set_debugreg(0UL, 1);
+	set_debugreg(0UL, 2);
+	set_debugreg(0UL, 3);
+	set_debugreg(0UL, 6);
+	set_debugreg(0UL, 7);
+
+	fpu_init(); 
+
+	raw_local_save_flags(kernel_eflags);
+}

arch/x86/kernel/signal_64.c

@@ -0,0 +1,495 @@
+/*
+ *  linux/arch/x86_64/kernel/signal.c
+ *
+ *  Copyright (C) 1991, 1992  Linus Torvalds
+ *  Copyright (C) 2000, 2001, 2002 Andi Kleen SuSE Labs
+ *
+ *  1997-11-28  Modified for POSIX.1b signals by Richard Henderson
+ *  2000-06-20  Pentium III FXSR, SSE support by Gareth Hughes
+ *  2000-2002   x86-64 support by Andi Kleen
+ */
+
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/kernel.h>
+#include <linux/signal.h>
+#include <linux/errno.h>
+#include <linux/wait.h>
+#include <linux/ptrace.h>
+#include <linux/unistd.h>
+#include <linux/stddef.h>
+#include <linux/personality.h>
+#include <linux/compiler.h>
+#include <asm/ucontext.h>
+#include <asm/uaccess.h>
+#include <asm/i387.h>
+#include <asm/proto.h>
+#include <asm/ia32_unistd.h>
+#include <asm/mce.h>
+
+/* #define DEBUG_SIG 1 */
+
+#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
+
+int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
+               sigset_t *set, struct pt_regs * regs); 
+int ia32_setup_frame(int sig, struct k_sigaction *ka,
+            sigset_t *set, struct pt_regs * regs); 
+
+asmlinkage long
+sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
+		struct pt_regs *regs)
+{
+	return do_sigaltstack(uss, uoss, regs->rsp);
+}
+
+
+/*
+ * Do a signal return; undo the signal stack.
+ */
+
+struct rt_sigframe
+{
+	char __user *pretcode;
+	struct ucontext uc;
+	struct siginfo info;
+};
+
+static int
+restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, unsigned long *prax)
+{
+	unsigned int err = 0;
+
+	/* Always make any pending restarted system calls return -EINTR */
+	current_thread_info()->restart_block.fn = do_no_restart_syscall;
+
+#define COPY(x)		err |= __get_user(regs->x, &sc->x)
+
+	COPY(rdi); COPY(rsi); COPY(rbp); COPY(rsp); COPY(rbx);
+	COPY(rdx); COPY(rcx); COPY(rip);
+	COPY(r8);
+	COPY(r9);
+	COPY(r10);
+	COPY(r11);
+	COPY(r12);
+	COPY(r13);
+	COPY(r14);
+	COPY(r15);
+
+	/* Kernel saves and restores only the CS segment register on signals,
+	 * which is the bare minimum needed to allow mixed 32/64-bit code.
+	 * App's signal handler can save/restore other segments if needed. */
+	{
+		unsigned cs;
+		err |= __get_user(cs, &sc->cs);
+		regs->cs = cs | 3;	/* Force into user mode */
+	}
+
+	{
+		unsigned int tmpflags;
+		err |= __get_user(tmpflags, &sc->eflags);
+		regs->eflags = (regs->eflags & ~0x40DD5) | (tmpflags & 0x40DD5);
+		regs->orig_rax = -1;		/* disable syscall checks */
+	}
+
+	{
+		struct _fpstate __user * buf;
+		err |= __get_user(buf, &sc->fpstate);
+
+		if (buf) {
+			if (!access_ok(VERIFY_READ, buf, sizeof(*buf)))
+				goto badframe;
+			err |= restore_i387(buf);
+		} else {
+			struct task_struct *me = current;
+			if (used_math()) {
+				clear_fpu(me);
+				clear_used_math();
+			}
+		}
+	}
+
+	err |= __get_user(*prax, &sc->rax);
+	return err;
+
+badframe:
+	return 1;
+}
+
+asmlinkage long sys_rt_sigreturn(struct pt_regs *regs)
+{
+	struct rt_sigframe __user *frame;
+	sigset_t set;
+	unsigned long eax;
+
+	frame = (struct rt_sigframe __user *)(regs->rsp - 8);
+	if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) {
+		goto badframe;
+	} 
+	if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) { 
+		goto badframe;
+	} 
+
+	sigdelsetmask(&set, ~_BLOCKABLE);
+	spin_lock_irq(&current->sighand->siglock);
+	current->blocked = set;
+	recalc_sigpending();
+	spin_unlock_irq(&current->sighand->siglock);
+	
+	if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &eax))
+		goto badframe;
+
+#ifdef DEBUG_SIG
+	printk("%d sigreturn rip:%lx rsp:%lx frame:%p rax:%lx\n",current->pid,regs->rip,regs->rsp,frame,eax);
+#endif
+
+	if (do_sigaltstack(&frame->uc.uc_stack, NULL, regs->rsp) == -EFAULT)
+		goto badframe;
+
+	return eax;
+
+badframe:
+	signal_fault(regs,frame,"sigreturn");
+	return 0;
+}	
+
+/*
+ * Set up a signal frame.
+ */
+
+static inline int
+setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, unsigned long mask, struct task_struct *me)
+{
+	int err = 0;
+
+	err |= __put_user(regs->cs, &sc->cs);
+	err |= __put_user(0, &sc->gs);
+	err |= __put_user(0, &sc->fs);
+
+	err |= __put_user(regs->rdi, &sc->rdi);
+	err |= __put_user(regs->rsi, &sc->rsi);
+	err |= __put_user(regs->rbp, &sc->rbp);
+	err |= __put_user(regs->rsp, &sc->rsp);
+	err |= __put_user(regs->rbx, &sc->rbx);
+	err |= __put_user(regs->rdx, &sc->rdx);
+	err |= __put_user(regs->rcx, &sc->rcx);
+	err |= __put_user(regs->rax, &sc->rax);
+	err |= __put_user(regs->r8, &sc->r8);
+	err |= __put_user(regs->r9, &sc->r9);
+	err |= __put_user(regs->r10, &sc->r10);
+	err |= __put_user(regs->r11, &sc->r11);
+	err |= __put_user(regs->r12, &sc->r12);
+	err |= __put_user(regs->r13, &sc->r13);
+	err |= __put_user(regs->r14, &sc->r14);
+	err |= __put_user(regs->r15, &sc->r15);
+	err |= __put_user(me->thread.trap_no, &sc->trapno);
+	err |= __put_user(me->thread.error_code, &sc->err);
+	err |= __put_user(regs->rip, &sc->rip);
+	err |= __put_user(regs->eflags, &sc->eflags);
+	err |= __put_user(mask, &sc->oldmask);
+	err |= __put_user(me->thread.cr2, &sc->cr2);
+
+	return err;
+}
+
+/*
+ * Determine which stack to use..
+ */
+
+static void __user *
+get_stack(struct k_sigaction *ka, struct pt_regs *regs, unsigned long size)
+{
+	unsigned long rsp;
+
+	/* Default to using normal stack - redzone*/
+	rsp = regs->rsp - 128;
+
+	/* This is the X/Open sanctioned signal stack switching.  */
+	if (ka->sa.sa_flags & SA_ONSTACK) {
+		if (sas_ss_flags(rsp) == 0)
+			rsp = current->sas_ss_sp + current->sas_ss_size;
+	}
+
+	return (void __user *)round_down(rsp - size, 16); 
+}
+
+static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
+			   sigset_t *set, struct pt_regs * regs)
+{
+	struct rt_sigframe __user *frame;
+	struct _fpstate __user *fp = NULL; 
+	int err = 0;
+	struct task_struct *me = current;
+
+	if (used_math()) {
+		fp = get_stack(ka, regs, sizeof(struct _fpstate)); 
+		frame = (void __user *)round_down(
+			(unsigned long)fp - sizeof(struct rt_sigframe), 16) - 8;
+
+		if (!access_ok(VERIFY_WRITE, fp, sizeof(struct _fpstate)))
+			goto give_sigsegv;
+
+		if (save_i387(fp) < 0) 
+			err |= -1; 
+	} else
+		frame = get_stack(ka, regs, sizeof(struct rt_sigframe)) - 8;
+
+	if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
+		goto give_sigsegv;
+
+	if (ka->sa.sa_flags & SA_SIGINFO) { 
+		err |= copy_siginfo_to_user(&frame->info, info);
+		if (err)
+			goto give_sigsegv;
+	}
+		
+	/* Create the ucontext.  */
+	err |= __put_user(0, &frame->uc.uc_flags);
+	err |= __put_user(0, &frame->uc.uc_link);
+	err |= __put_user(me->sas_ss_sp, &frame->uc.uc_stack.ss_sp);
+	err |= __put_user(sas_ss_flags(regs->rsp),
+			  &frame->uc.uc_stack.ss_flags);
+	err |= __put_user(me->sas_ss_size, &frame->uc.uc_stack.ss_size);
+	err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, set->sig[0], me);
+	err |= __put_user(fp, &frame->uc.uc_mcontext.fpstate);
+	if (sizeof(*set) == 16) { 
+		__put_user(set->sig[0], &frame->uc.uc_sigmask.sig[0]);
+		__put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]); 
+	} else
+		err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
+
+	/* Set up to return from userspace.  If provided, use a stub
+	   already in userspace.  */
+	/* x86-64 should always use SA_RESTORER. */
+	if (ka->sa.sa_flags & SA_RESTORER) {
+		err |= __put_user(ka->sa.sa_restorer, &frame->pretcode);
+	} else {
+		/* could use a vstub here */
+		goto give_sigsegv; 
+	}
+
+	if (err)
+		goto give_sigsegv;
+
+#ifdef DEBUG_SIG
+	printk("%d old rip %lx old rsp %lx old rax %lx\n", current->pid,regs->rip,regs->rsp,regs->rax);
+#endif
+
+	/* Set up registers for signal handler */
+	regs->rdi = sig;
+	/* In case the signal handler was declared without prototypes */ 
+	regs->rax = 0;	
+
+	/* This also works for non SA_SIGINFO handlers because they expect the
+	   next argument after the signal number on the stack. */
+	regs->rsi = (unsigned long)&frame->info; 
+	regs->rdx = (unsigned long)&frame->uc; 
+	regs->rip = (unsigned long) ka->sa.sa_handler;
+
+	regs->rsp = (unsigned long)frame;
+
+	/* Set up the CS register to run signal handlers in 64-bit mode,
+	   even if the handler happens to be interrupting 32-bit code. */
+	regs->cs = __USER_CS;
+
+	/* This, by contrast, has nothing to do with segment registers -
+	   see include/asm-x86_64/uaccess.h for details. */
+	set_fs(USER_DS);
+
+	regs->eflags &= ~TF_MASK;
+	if (test_thread_flag(TIF_SINGLESTEP))
+		ptrace_notify(SIGTRAP);
+#ifdef DEBUG_SIG
+	printk("SIG deliver (%s:%d): sp=%p pc=%lx ra=%p\n",
+		current->comm, current->pid, frame, regs->rip, frame->pretcode);
+#endif
+
+	return 0;
+
+give_sigsegv:
+	force_sigsegv(sig, current);
+	return -EFAULT;
+}
+
+/*
+ * OK, we're invoking a handler
+ */	
+
+static int
+handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
+		sigset_t *oldset, struct pt_regs *regs)
+{
+	int ret;
+
+#ifdef DEBUG_SIG
+	printk("handle_signal pid:%d sig:%lu rip:%lx rsp:%lx regs=%p\n",
+		current->pid, sig,
+		regs->rip, regs->rsp, regs);
+#endif
+
+	/* Are we from a system call? */
+	if ((long)regs->orig_rax >= 0) {
+		/* If so, check system call restarting.. */
+		switch (regs->rax) {
+		        case -ERESTART_RESTARTBLOCK:
+			case -ERESTARTNOHAND:
+				regs->rax = -EINTR;
+				break;
+
+			case -ERESTARTSYS:
+				if (!(ka->sa.sa_flags & SA_RESTART)) {
+					regs->rax = -EINTR;
+					break;
+				}
+				/* fallthrough */
+			case -ERESTARTNOINTR:
+				regs->rax = regs->orig_rax;
+				regs->rip -= 2;
+				break;
+		}
+	}
+
+	/*
+	 * If TF is set due to a debugger (PT_DTRACE), clear the TF
+	 * flag so that register information in the sigcontext is
+	 * correct.
+	 */
+	if (unlikely(regs->eflags & TF_MASK)) {
+		if (likely(current->ptrace & PT_DTRACE)) {
+			current->ptrace &= ~PT_DTRACE;
+			regs->eflags &= ~TF_MASK;
+		}
+	}
+
+#ifdef CONFIG_IA32_EMULATION
+	if (test_thread_flag(TIF_IA32)) {
+		if (ka->sa.sa_flags & SA_SIGINFO)
+			ret = ia32_setup_rt_frame(sig, ka, info, oldset, regs);
+		else
+			ret = ia32_setup_frame(sig, ka, oldset, regs);
+	} else 
+#endif
+	ret = setup_rt_frame(sig, ka, info, oldset, regs);
+
+	if (ret == 0) {
+		spin_lock_irq(&current->sighand->siglock);
+		sigorsets(&current->blocked,&current->blocked,&ka->sa.sa_mask);
+		if (!(ka->sa.sa_flags & SA_NODEFER))
+			sigaddset(&current->blocked,sig);
+		recalc_sigpending();
+		spin_unlock_irq(&current->sighand->siglock);
+	}
+
+	return ret;
+}
+
+/*
+ * Note that 'init' is a special process: it doesn't get signals it doesn't
+ * want to handle. Thus you cannot kill init even with a SIGKILL even by
+ * mistake.
+ */
+static void do_signal(struct pt_regs *regs)
+{
+	struct k_sigaction ka;
+	siginfo_t info;
+	int signr;
+	sigset_t *oldset;
+
+	/*
+	 * We want the common case to go fast, which
+	 * is why we may in certain cases get here from
+	 * kernel mode. Just return without doing anything
+	 * if so.
+	 */
+	if (!user_mode(regs))
+		return;
+
+	if (test_thread_flag(TIF_RESTORE_SIGMASK))
+		oldset = &current->saved_sigmask;
+	else
+		oldset = &current->blocked;
+
+	signr = get_signal_to_deliver(&info, &ka, regs, NULL);
+	if (signr > 0) {
+		/* Reenable any watchpoints before delivering the
+		 * signal to user space. The processor register will
+		 * have been cleared if the watchpoint triggered
+		 * inside the kernel.
+		 */
+		if (current->thread.debugreg7)
+			set_debugreg(current->thread.debugreg7, 7);
+
+		/* Whee!  Actually deliver the signal.  */
+		if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
+			/* a signal was successfully delivered; the saved
+			 * sigmask will have been stored in the signal frame,
+			 * and will be restored by sigreturn, so we can simply
+			 * clear the TIF_RESTORE_SIGMASK flag */
+			clear_thread_flag(TIF_RESTORE_SIGMASK);
+		}
+		return;
+	}
+
+	/* Did we come from a system call? */
+	if ((long)regs->orig_rax >= 0) {
+		/* Restart the system call - no handlers present */
+		long res = regs->rax;
+		switch (res) {
+		case -ERESTARTNOHAND:
+		case -ERESTARTSYS:
+		case -ERESTARTNOINTR:
+			regs->rax = regs->orig_rax;
+			regs->rip -= 2;
+			break;
+		case -ERESTART_RESTARTBLOCK:
+			regs->rax = test_thread_flag(TIF_IA32) ?
+					__NR_ia32_restart_syscall :
+					__NR_restart_syscall;
+			regs->rip -= 2;
+			break;
+		}
+	}
+
+	/* if there's no signal to deliver, we just put the saved sigmask
+	   back. */
+	if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
+		clear_thread_flag(TIF_RESTORE_SIGMASK);
+		sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
+	}
+}
+
+void
+do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
+{
+#ifdef DEBUG_SIG
+	printk("do_notify_resume flags:%x rip:%lx rsp:%lx caller:%p pending:%x\n",
+	       thread_info_flags, regs->rip, regs->rsp, __builtin_return_address(0),signal_pending(current)); 
+#endif
+	       
+	/* Pending single-step? */
+	if (thread_info_flags & _TIF_SINGLESTEP) {
+		regs->eflags |= TF_MASK;
+		clear_thread_flag(TIF_SINGLESTEP);
+	}
+
+#ifdef CONFIG_X86_MCE
+	/* notify userspace of pending MCEs */
+	if (thread_info_flags & _TIF_MCE_NOTIFY)
+		mce_notify_user();
+#endif /* CONFIG_X86_MCE */
+
+	/* deal with pending signal delivery */
+	if (thread_info_flags & (_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK))
+		do_signal(regs);
+}
+
+void signal_fault(struct pt_regs *regs, void __user *frame, char *where)
+{ 
+	struct task_struct *me = current; 
+	if (show_unhandled_signals && printk_ratelimit())
+		printk("%s[%d] bad frame in %s frame:%p rip:%lx rsp:%lx orax:%lx\n",
+	       me->comm,me->pid,where,frame,regs->rip,regs->rsp,regs->orig_rax); 
+
+	force_sig(SIGSEGV, me); 
+} 

arch/x86/kernel/smp_64.c

@@ -0,0 +1,523 @@
+/*
+ *	Intel SMP support routines.
+ *
+ *	(c) 1995 Alan Cox, Building #3 <alan@redhat.com>
+ *	(c) 1998-99, 2000 Ingo Molnar <mingo@redhat.com>
+ *      (c) 2002,2003 Andi Kleen, SuSE Labs.
+ *
+ *	This code is released under the GNU General Public License version 2 or
+ *	later.
+ */
+
+#include <linux/init.h>
+
+#include <linux/mm.h>
+#include <linux/delay.h>
+#include <linux/spinlock.h>
+#include <linux/smp.h>
+#include <linux/kernel_stat.h>
+#include <linux/mc146818rtc.h>
+#include <linux/interrupt.h>
+
+#include <asm/mtrr.h>
+#include <asm/pgalloc.h>
+#include <asm/tlbflush.h>
+#include <asm/mach_apic.h>
+#include <asm/mmu_context.h>
+#include <asm/proto.h>
+#include <asm/apicdef.h>
+#include <asm/idle.h>
+
+/*
+ *	Smarter SMP flushing macros. 
+ *		c/o Linus Torvalds.
+ *
+ *	These mean you can really definitely utterly forget about
+ *	writing to user space from interrupts. (Its not allowed anyway).
+ *
+ *	Optimizations Manfred Spraul <manfred@colorfullife.com>
+ *
+ * 	More scalable flush, from Andi Kleen
+ *
+ * 	To avoid global state use 8 different call vectors.
+ * 	Each CPU uses a specific vector to trigger flushes on other
+ * 	CPUs. Depending on the received vector the target CPUs look into
+ *	the right per cpu variable for the flush data.
+ *
+ * 	With more than 8 CPUs they are hashed to the 8 available
+ * 	vectors. The limited global vector space forces us to this right now.
+ *	In future when interrupts are split into per CPU domains this could be
+ *	fixed, at the cost of triggering multiple IPIs in some cases.
+ */
+
+union smp_flush_state {
+	struct {
+		cpumask_t flush_cpumask;
+		struct mm_struct *flush_mm;
+		unsigned long flush_va;
+#define FLUSH_ALL	-1ULL
+		spinlock_t tlbstate_lock;
+	};
+	char pad[SMP_CACHE_BYTES];
+} ____cacheline_aligned;
+
+/* State is put into the per CPU data section, but padded
+   to a full cache line because other CPUs can access it and we don't
+   want false sharing in the per cpu data segment. */
+static DEFINE_PER_CPU(union smp_flush_state, flush_state);
+
+/*
+ * We cannot call mmdrop() because we are in interrupt context, 
+ * instead update mm->cpu_vm_mask.
+ */
+static inline void leave_mm(int cpu)
+{
+	if (read_pda(mmu_state) == TLBSTATE_OK)
+		BUG();
+	cpu_clear(cpu, read_pda(active_mm)->cpu_vm_mask);
+	load_cr3(swapper_pg_dir);
+}
+
+/*
+ *
+ * The flush IPI assumes that a thread switch happens in this order:
+ * [cpu0: the cpu that switches]
+ * 1) switch_mm() either 1a) or 1b)
+ * 1a) thread switch to a different mm
+ * 1a1) cpu_clear(cpu, old_mm->cpu_vm_mask);
+ * 	Stop ipi delivery for the old mm. This is not synchronized with
+ * 	the other cpus, but smp_invalidate_interrupt ignore flush ipis
+ * 	for the wrong mm, and in the worst case we perform a superfluous
+ * 	tlb flush.
+ * 1a2) set cpu mmu_state to TLBSTATE_OK
+ * 	Now the smp_invalidate_interrupt won't call leave_mm if cpu0
+ *	was in lazy tlb mode.
+ * 1a3) update cpu active_mm
+ * 	Now cpu0 accepts tlb flushes for the new mm.
+ * 1a4) cpu_set(cpu, new_mm->cpu_vm_mask);
+ * 	Now the other cpus will send tlb flush ipis.
+ * 1a4) change cr3.
+ * 1b) thread switch without mm change
+ *	cpu active_mm is correct, cpu0 already handles
+ *	flush ipis.
+ * 1b1) set cpu mmu_state to TLBSTATE_OK
+ * 1b2) test_and_set the cpu bit in cpu_vm_mask.
+ * 	Atomically set the bit [other cpus will start sending flush ipis],
+ * 	and test the bit.
+ * 1b3) if the bit was 0: leave_mm was called, flush the tlb.
+ * 2) switch %%esp, ie current
+ *
+ * The interrupt must handle 2 special cases:
+ * - cr3 is changed before %%esp, ie. it cannot use current->{active_,}mm.
+ * - the cpu performs speculative tlb reads, i.e. even if the cpu only
+ *   runs in kernel space, the cpu could load tlb entries for user space
+ *   pages.
+ *
+ * The good news is that cpu mmu_state is local to each cpu, no
+ * write/read ordering problems.
+ */
+
+/*
+ * TLB flush IPI:
+ *
+ * 1) Flush the tlb entries if the cpu uses the mm that's being flushed.
+ * 2) Leave the mm if we are in the lazy tlb mode.
+ *
+ * Interrupts are disabled.
+ */
+
+asmlinkage void smp_invalidate_interrupt(struct pt_regs *regs)
+{
+	int cpu;
+	int sender;
+	union smp_flush_state *f;
+
+	cpu = smp_processor_id();
+	/*
+	 * orig_rax contains the negated interrupt vector.
+	 * Use that to determine where the sender put the data.
+	 */
+	sender = ~regs->orig_rax - INVALIDATE_TLB_VECTOR_START;
+	f = &per_cpu(flush_state, sender);
+
+	if (!cpu_isset(cpu, f->flush_cpumask))
+		goto out;
+		/* 
+		 * This was a BUG() but until someone can quote me the
+		 * line from the intel manual that guarantees an IPI to
+		 * multiple CPUs is retried _only_ on the erroring CPUs
+		 * its staying as a return
+		 *
+		 * BUG();
+		 */
+		 
+	if (f->flush_mm == read_pda(active_mm)) {
+		if (read_pda(mmu_state) == TLBSTATE_OK) {
+			if (f->flush_va == FLUSH_ALL)
+				local_flush_tlb();
+			else
+				__flush_tlb_one(f->flush_va);
+		} else
+			leave_mm(cpu);
+	}
+out:
+	ack_APIC_irq();
+	cpu_clear(cpu, f->flush_cpumask);
+}
+
+static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
+						unsigned long va)
+{
+	int sender;
+	union smp_flush_state *f;
+
+	/* Caller has disabled preemption */
+	sender = smp_processor_id() % NUM_INVALIDATE_TLB_VECTORS;
+	f = &per_cpu(flush_state, sender);
+
+	/* Could avoid this lock when
+	   num_online_cpus() <= NUM_INVALIDATE_TLB_VECTORS, but it is
+	   probably not worth checking this for a cache-hot lock. */
+	spin_lock(&f->tlbstate_lock);
+
+	f->flush_mm = mm;
+	f->flush_va = va;
+	cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask);
+
+	/*
+	 * We have to send the IPI only to
+	 * CPUs affected.
+	 */
+	send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR_START + sender);
+
+	while (!cpus_empty(f->flush_cpumask))
+		cpu_relax();
+
+	f->flush_mm = NULL;
+	f->flush_va = 0;
+	spin_unlock(&f->tlbstate_lock);
+}
+
+int __cpuinit init_smp_flush(void)
+{
+	int i;
+	for_each_cpu_mask(i, cpu_possible_map) {
+		spin_lock_init(&per_cpu(flush_state, i).tlbstate_lock);
+	}
+	return 0;
+}
+
+core_initcall(init_smp_flush);
+	
+void flush_tlb_current_task(void)
+{
+	struct mm_struct *mm = current->mm;
+	cpumask_t cpu_mask;
+
+	preempt_disable();
+	cpu_mask = mm->cpu_vm_mask;
+	cpu_clear(smp_processor_id(), cpu_mask);
+
+	local_flush_tlb();
+	if (!cpus_empty(cpu_mask))
+		flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
+	preempt_enable();
+}
+EXPORT_SYMBOL(flush_tlb_current_task);
+
+void flush_tlb_mm (struct mm_struct * mm)
+{
+	cpumask_t cpu_mask;
+
+	preempt_disable();
+	cpu_mask = mm->cpu_vm_mask;
+	cpu_clear(smp_processor_id(), cpu_mask);
+
+	if (current->active_mm == mm) {
+		if (current->mm)
+			local_flush_tlb();
+		else
+			leave_mm(smp_processor_id());
+	}
+	if (!cpus_empty(cpu_mask))
+		flush_tlb_others(cpu_mask, mm, FLUSH_ALL);
+
+	preempt_enable();
+}
+EXPORT_SYMBOL(flush_tlb_mm);
+
+void flush_tlb_page(struct vm_area_struct * vma, unsigned long va)
+{
+	struct mm_struct *mm = vma->vm_mm;
+	cpumask_t cpu_mask;
+
+	preempt_disable();
+	cpu_mask = mm->cpu_vm_mask;
+	cpu_clear(smp_processor_id(), cpu_mask);
+
+	if (current->active_mm == mm) {
+		if(current->mm)
+			__flush_tlb_one(va);
+		 else
+		 	leave_mm(smp_processor_id());
+	}
+
+	if (!cpus_empty(cpu_mask))
+		flush_tlb_others(cpu_mask, mm, va);
+
+	preempt_enable();
+}
+EXPORT_SYMBOL(flush_tlb_page);
+
+static void do_flush_tlb_all(void* info)
+{
+	unsigned long cpu = smp_processor_id();
+
+	__flush_tlb_all();
+	if (read_pda(mmu_state) == TLBSTATE_LAZY)
+		leave_mm(cpu);
+}
+
+void flush_tlb_all(void)
+{
+	on_each_cpu(do_flush_tlb_all, NULL, 1, 1);
+}
+
+/*
+ * this function sends a 'reschedule' IPI to another CPU.
+ * it goes straight through and wastes no time serializing
+ * anything. Worst case is that we lose a reschedule ...
+ */
+
+void smp_send_reschedule(int cpu)
+{
+	send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
+}
+
+/*
+ * Structure and data for smp_call_function(). This is designed to minimise
+ * static memory requirements. It also looks cleaner.
+ */
+static DEFINE_SPINLOCK(call_lock);
+
+struct call_data_struct {
+	void (*func) (void *info);
+	void *info;
+	atomic_t started;
+	atomic_t finished;
+	int wait;
+};
+
+static struct call_data_struct * call_data;
+
+void lock_ipi_call_lock(void)
+{
+	spin_lock_irq(&call_lock);
+}
+
+void unlock_ipi_call_lock(void)
+{
+	spin_unlock_irq(&call_lock);
+}
+
+/*
+ * this function sends a 'generic call function' IPI to one other CPU
+ * in the system.
+ *
+ * cpu is a standard Linux logical CPU number.
+ */
+static void
+__smp_call_function_single(int cpu, void (*func) (void *info), void *info,
+				int nonatomic, int wait)
+{
+	struct call_data_struct data;
+	int cpus = 1;
+
+	data.func = func;
+	data.info = info;
+	atomic_set(&data.started, 0);
+	data.wait = wait;
+	if (wait)
+		atomic_set(&data.finished, 0);
+
+	call_data = &data;
+	wmb();
+	/* Send a message to all other CPUs and wait for them to respond */
+	send_IPI_mask(cpumask_of_cpu(cpu), CALL_FUNCTION_VECTOR);
+
+	/* Wait for response */
+	while (atomic_read(&data.started) != cpus)
+		cpu_relax();
+
+	if (!wait)
+		return;
+
+	while (atomic_read(&data.finished) != cpus)
+		cpu_relax();
+}
+
+/*
+ * smp_call_function_single - Run a function on a specific CPU
+ * @func: The function to run. This must be fast and non-blocking.
+ * @info: An arbitrary pointer to pass to the function.
+ * @nonatomic: Currently unused.
+ * @wait: If true, wait until function has completed on other CPUs.
+ *
+ * Retrurns 0 on success, else a negative status code.
+ *
+ * Does not return until the remote CPU is nearly ready to execute <func>
+ * or is or has executed.
+ */
+
+int smp_call_function_single (int cpu, void (*func) (void *info), void *info,
+	int nonatomic, int wait)
+{
+	/* prevent preemption and reschedule on another processor */
+	int me = get_cpu();
+
+	/* Can deadlock when called with interrupts disabled */
+	WARN_ON(irqs_disabled());
+
+	if (cpu == me) {
+		local_irq_disable();
+		func(info);
+		local_irq_enable();
+		put_cpu();
+		return 0;
+	}
+
+	spin_lock(&call_lock);
+	__smp_call_function_single(cpu, func, info, nonatomic, wait);
+	spin_unlock(&call_lock);
+	put_cpu();
+	return 0;
+}
+EXPORT_SYMBOL(smp_call_function_single);
+
+/*
+ * this function sends a 'generic call function' IPI to all other CPUs
+ * in the system.
+ */
+static void __smp_call_function (void (*func) (void *info), void *info,
+				int nonatomic, int wait)
+{
+	struct call_data_struct data;
+	int cpus = num_online_cpus()-1;
+
+	if (!cpus)
+		return;
+
+	data.func = func;
+	data.info = info;
+	atomic_set(&data.started, 0);
+	data.wait = wait;
+	if (wait)
+		atomic_set(&data.finished, 0);
+
+	call_data = &data;
+	wmb();
+	/* Send a message to all other CPUs and wait for them to respond */
+	send_IPI_allbutself(CALL_FUNCTION_VECTOR);
+
+	/* Wait for response */
+	while (atomic_read(&data.started) != cpus)
+		cpu_relax();
+
+	if (!wait)
+		return;
+
+	while (atomic_read(&data.finished) != cpus)
+		cpu_relax();
+}
+
+/*
+ * smp_call_function - run a function on all other CPUs.
+ * @func: The function to run. This must be fast and non-blocking.
+ * @info: An arbitrary pointer to pass to the function.
+ * @nonatomic: currently unused.
+ * @wait: If true, wait (atomically) until function has completed on other
+ *        CPUs.
+ *
+ * Returns 0 on success, else a negative status code. Does not return until
+ * remote CPUs are nearly ready to execute func or are or have executed.
+ *
+ * You must not call this function with disabled interrupts or from a
+ * hardware interrupt handler or from a bottom half handler.
+ * Actually there are a few legal cases, like panic.
+ */
+int smp_call_function (void (*func) (void *info), void *info, int nonatomic,
+			int wait)
+{
+	spin_lock(&call_lock);
+	__smp_call_function(func,info,nonatomic,wait);
+	spin_unlock(&call_lock);
+	return 0;
+}
+EXPORT_SYMBOL(smp_call_function);
+
+static void stop_this_cpu(void *dummy)
+{
+	local_irq_disable();
+	/*
+	 * Remove this CPU:
+	 */
+	cpu_clear(smp_processor_id(), cpu_online_map);
+	disable_local_APIC();
+	for (;;) 
+		halt();
+} 
+
+void smp_send_stop(void)
+{
+	int nolock;
+	unsigned long flags;
+
+	if (reboot_force)
+		return;
+
+	/* Don't deadlock on the call lock in panic */
+	nolock = !spin_trylock(&call_lock);
+	local_irq_save(flags);
+	__smp_call_function(stop_this_cpu, NULL, 0, 0);
+	if (!nolock)
+		spin_unlock(&call_lock);
+	disable_local_APIC();
+	local_irq_restore(flags);
+}
+
+/*
+ * Reschedule call back. Nothing to do,
+ * all the work is done automatically when
+ * we return from the interrupt.
+ */
+asmlinkage void smp_reschedule_interrupt(void)
+{
+	ack_APIC_irq();
+}
+
+asmlinkage void smp_call_function_interrupt(void)
+{
+	void (*func) (void *info) = call_data->func;
+	void *info = call_data->info;
+	int wait = call_data->wait;
+
+	ack_APIC_irq();
+	/*
+	 * Notify initiating CPU that I've grabbed the data and am
+	 * about to execute the function
+	 */
+	mb();
+	atomic_inc(&call_data->started);
+	/*
+	 * At this point the info structure may be out of scope unless wait==1
+	 */
+	exit_idle();
+	irq_enter();
+	(*func)(info);
+	irq_exit();
+	if (wait) {
+		mb();
+		atomic_inc(&call_data->finished);
+	}
+}
+

arch/x86/kernel/stacktrace.c

@@ -0,0 +1,54 @@
+/*
+ * arch/x86_64/kernel/stacktrace.c
+ *
+ * Stack trace management functions
+ *
+ *  Copyright (C) 2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
+ */
+#include <linux/sched.h>
+#include <linux/stacktrace.h>
+#include <linux/module.h>
+#include <asm/stacktrace.h>
+
+static void save_stack_warning(void *data, char *msg)
+{
+}
+
+static void
+save_stack_warning_symbol(void *data, char *msg, unsigned long symbol)
+{
+}
+
+static int save_stack_stack(void *data, char *name)
+{
+	return -1;
+}
+
+static void save_stack_address(void *data, unsigned long addr)
+{
+	struct stack_trace *trace = (struct stack_trace *)data;
+	if (trace->skip > 0) {
+		trace->skip--;
+		return;
+	}
+	if (trace->nr_entries < trace->max_entries)
+		trace->entries[trace->nr_entries++] = addr;
+}
+
+static struct stacktrace_ops save_stack_ops = {
+	.warning = save_stack_warning,
+	.warning_symbol = save_stack_warning_symbol,
+	.stack = save_stack_stack,
+	.address = save_stack_address,
+};
+
+/*
+ * Save stack-backtrace addresses into a stack_trace buffer.
+ */
+void save_stack_trace(struct stack_trace *trace)
+{
+	dump_trace(current, NULL, NULL, &save_stack_ops, trace);
+	if (trace->nr_entries < trace->max_entries)
+		trace->entries[trace->nr_entries++] = ULONG_MAX;
+}
+EXPORT_SYMBOL(save_stack_trace);

arch/x86/kernel/suspend_64.c

@@ -0,0 +1,239 @@
+/*
+ * Suspend support specific for i386.
+ *
+ * Distribute under GPLv2
+ *
+ * Copyright (c) 2002 Pavel Machek <pavel@suse.cz>
+ * Copyright (c) 2001 Patrick Mochel <mochel@osdl.org>
+ */
+
+#include <linux/smp.h>
+#include <linux/suspend.h>
+#include <asm/proto.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/mtrr.h>
+
+/* References to section boundaries */
+extern const void __nosave_begin, __nosave_end;
+
+struct saved_context saved_context;
+
+unsigned long saved_context_eax, saved_context_ebx, saved_context_ecx, saved_context_edx;
+unsigned long saved_context_esp, saved_context_ebp, saved_context_esi, saved_context_edi;
+unsigned long saved_context_r08, saved_context_r09, saved_context_r10, saved_context_r11;
+unsigned long saved_context_r12, saved_context_r13, saved_context_r14, saved_context_r15;
+unsigned long saved_context_eflags;
+
+void __save_processor_state(struct saved_context *ctxt)
+{
+	kernel_fpu_begin();
+
+	/*
+	 * descriptor tables
+	 */
+	asm volatile ("sgdt %0" : "=m" (ctxt->gdt_limit));
+	asm volatile ("sidt %0" : "=m" (ctxt->idt_limit));
+	asm volatile ("str %0"  : "=m" (ctxt->tr));
+
+	/* XMM0..XMM15 should be handled by kernel_fpu_begin(). */
+	/*
+	 * segment registers
+	 */
+	asm volatile ("movw %%ds, %0" : "=m" (ctxt->ds));
+	asm volatile ("movw %%es, %0" : "=m" (ctxt->es));
+	asm volatile ("movw %%fs, %0" : "=m" (ctxt->fs));
+	asm volatile ("movw %%gs, %0" : "=m" (ctxt->gs));
+	asm volatile ("movw %%ss, %0" : "=m" (ctxt->ss));
+
+	rdmsrl(MSR_FS_BASE, ctxt->fs_base);
+	rdmsrl(MSR_GS_BASE, ctxt->gs_base);
+	rdmsrl(MSR_KERNEL_GS_BASE, ctxt->gs_kernel_base);
+	mtrr_save_fixed_ranges(NULL);
+
+	/*
+	 * control registers 
+	 */
+	rdmsrl(MSR_EFER, ctxt->efer);
+	ctxt->cr0 = read_cr0();
+	ctxt->cr2 = read_cr2();
+	ctxt->cr3 = read_cr3();
+	ctxt->cr4 = read_cr4();
+	ctxt->cr8 = read_cr8();
+}
+
+void save_processor_state(void)
+{
+	__save_processor_state(&saved_context);
+}
+
+static void do_fpu_end(void)
+{
+	/*
+	 * Restore FPU regs if necessary
+	 */
+	kernel_fpu_end();
+}
+
+void __restore_processor_state(struct saved_context *ctxt)
+{
+	/*
+	 * control registers
+	 */
+	wrmsrl(MSR_EFER, ctxt->efer);
+	write_cr8(ctxt->cr8);
+	write_cr4(ctxt->cr4);
+	write_cr3(ctxt->cr3);
+	write_cr2(ctxt->cr2);
+	write_cr0(ctxt->cr0);
+
+	/*
+	 * now restore the descriptor tables to their proper values
+	 * ltr is done i fix_processor_context().
+	 */
+	asm volatile ("lgdt %0" :: "m" (ctxt->gdt_limit));
+	asm volatile ("lidt %0" :: "m" (ctxt->idt_limit));
+
+	/*
+	 * segment registers
+	 */
+	asm volatile ("movw %0, %%ds" :: "r" (ctxt->ds));
+	asm volatile ("movw %0, %%es" :: "r" (ctxt->es));
+	asm volatile ("movw %0, %%fs" :: "r" (ctxt->fs));
+	load_gs_index(ctxt->gs);
+	asm volatile ("movw %0, %%ss" :: "r" (ctxt->ss));
+
+	wrmsrl(MSR_FS_BASE, ctxt->fs_base);
+	wrmsrl(MSR_GS_BASE, ctxt->gs_base);
+	wrmsrl(MSR_KERNEL_GS_BASE, ctxt->gs_kernel_base);
+
+	fix_processor_context();
+
+	do_fpu_end();
+	mtrr_ap_init();
+}
+
+void restore_processor_state(void)
+{
+	__restore_processor_state(&saved_context);
+}
+
+void fix_processor_context(void)
+{
+	int cpu = smp_processor_id();
+	struct tss_struct *t = &per_cpu(init_tss, cpu);
+
+	set_tss_desc(cpu,t);	/* This just modifies memory; should not be neccessary. But... This is neccessary, because 386 hardware has concept of busy TSS or some similar stupidity. */
+
+	cpu_gdt(cpu)[GDT_ENTRY_TSS].type = 9;
+
+	syscall_init();                         /* This sets MSR_*STAR and related */
+	load_TR_desc();				/* This does ltr */
+	load_LDT(&current->active_mm->context);	/* This does lldt */
+
+	/*
+	 * Now maybe reload the debug registers
+	 */
+	if (current->thread.debugreg7){
+                loaddebug(&current->thread, 0);
+                loaddebug(&current->thread, 1);
+                loaddebug(&current->thread, 2);
+                loaddebug(&current->thread, 3);
+                /* no 4 and 5 */
+                loaddebug(&current->thread, 6);
+                loaddebug(&current->thread, 7);
+	}
+
+}
+
+#ifdef CONFIG_HIBERNATION
+/* Defined in arch/x86_64/kernel/suspend_asm.S */
+extern int restore_image(void);
+
+pgd_t *temp_level4_pgt;
+
+static int res_phys_pud_init(pud_t *pud, unsigned long address, unsigned long end)
+{
+	long i, j;
+
+	i = pud_index(address);
+	pud = pud + i;
+	for (; i < PTRS_PER_PUD; pud++, i++) {
+		unsigned long paddr;
+		pmd_t *pmd;
+
+		paddr = address + i*PUD_SIZE;
+		if (paddr >= end)
+			break;
+
+		pmd = (pmd_t *)get_safe_page(GFP_ATOMIC);
+		if (!pmd)
+			return -ENOMEM;
+		set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE));
+		for (j = 0; j < PTRS_PER_PMD; pmd++, j++, paddr += PMD_SIZE) {
+			unsigned long pe;
+
+			if (paddr >= end)
+				break;
+			pe = _PAGE_NX | _PAGE_PSE | _KERNPG_TABLE | paddr;
+			pe &= __supported_pte_mask;
+			set_pmd(pmd, __pmd(pe));
+		}
+	}
+	return 0;
+}
+
+static int set_up_temporary_mappings(void)
+{
+	unsigned long start, end, next;
+	int error;
+
+	temp_level4_pgt = (pgd_t *)get_safe_page(GFP_ATOMIC);
+	if (!temp_level4_pgt)
+		return -ENOMEM;
+
+	/* It is safe to reuse the original kernel mapping */
+	set_pgd(temp_level4_pgt + pgd_index(__START_KERNEL_map),
+		init_level4_pgt[pgd_index(__START_KERNEL_map)]);
+
+	/* Set up the direct mapping from scratch */
+	start = (unsigned long)pfn_to_kaddr(0);
+	end = (unsigned long)pfn_to_kaddr(end_pfn);
+
+	for (; start < end; start = next) {
+		pud_t *pud = (pud_t *)get_safe_page(GFP_ATOMIC);
+		if (!pud)
+			return -ENOMEM;
+		next = start + PGDIR_SIZE;
+		if (next > end)
+			next = end;
+		if ((error = res_phys_pud_init(pud, __pa(start), __pa(next))))
+			return error;
+		set_pgd(temp_level4_pgt + pgd_index(start),
+			mk_kernel_pgd(__pa(pud)));
+	}
+	return 0;
+}
+
+int swsusp_arch_resume(void)
+{
+	int error;
+
+	/* We have got enough memory and from now on we cannot recover */
+	if ((error = set_up_temporary_mappings()))
+		return error;
+	restore_image();
+	return 0;
+}
+
+/*
+ *	pfn_is_nosave - check if given pfn is in the 'nosave' section
+ */
+
+int pfn_is_nosave(unsigned long pfn)
+{
+	unsigned long nosave_begin_pfn = __pa_symbol(&__nosave_begin) >> PAGE_SHIFT;
+	unsigned long nosave_end_pfn = PAGE_ALIGN(__pa_symbol(&__nosave_end)) >> PAGE_SHIFT;
+	return (pfn >= nosave_begin_pfn) && (pfn < nosave_end_pfn);
+}
+#endif /* CONFIG_HIBERNATION */

arch/x86/kernel/suspend_asm_64.S

@@ -0,0 +1,110 @@
+/* Copyright 2004,2005 Pavel Machek <pavel@suse.cz>, Andi Kleen <ak@suse.de>, Rafael J. Wysocki <rjw@sisk.pl>
+ *
+ * Distribute under GPLv2.
+ *
+ * swsusp_arch_resume may not use any stack, nor any variable that is
+ * not "NoSave" during copying pages:
+ *
+ * Its rewriting one kernel image with another. What is stack in "old"
+ * image could very well be data page in "new" image, and overwriting
+ * your own stack under you is bad idea.
+ */
+	
+	.text
+#include <linux/linkage.h>
+#include <asm/segment.h>
+#include <asm/page.h>
+#include <asm/asm-offsets.h>
+
+ENTRY(swsusp_arch_suspend)
+
+	movq %rsp, saved_context_esp(%rip)
+	movq %rax, saved_context_eax(%rip)
+	movq %rbx, saved_context_ebx(%rip)
+	movq %rcx, saved_context_ecx(%rip)
+	movq %rdx, saved_context_edx(%rip)
+	movq %rbp, saved_context_ebp(%rip)
+	movq %rsi, saved_context_esi(%rip)
+	movq %rdi, saved_context_edi(%rip)
+	movq %r8,  saved_context_r08(%rip)
+	movq %r9,  saved_context_r09(%rip)
+	movq %r10, saved_context_r10(%rip)
+	movq %r11, saved_context_r11(%rip)
+	movq %r12, saved_context_r12(%rip)
+	movq %r13, saved_context_r13(%rip)
+	movq %r14, saved_context_r14(%rip)
+	movq %r15, saved_context_r15(%rip)
+	pushfq ; popq saved_context_eflags(%rip)
+
+	call swsusp_save
+	ret
+
+ENTRY(restore_image)
+	/* switch to temporary page tables */
+	movq	$__PAGE_OFFSET, %rdx
+	movq	temp_level4_pgt(%rip), %rax
+	subq	%rdx, %rax
+	movq	%rax, %cr3
+	/* Flush TLB */
+	movq	mmu_cr4_features(%rip), %rax
+	movq	%rax, %rdx
+	andq	$~(1<<7), %rdx	# PGE
+	movq	%rdx, %cr4;  # turn off PGE
+	movq	%cr3, %rcx;  # flush TLB
+	movq	%rcx, %cr3;
+	movq	%rax, %cr4;  # turn PGE back on
+
+	movq	restore_pblist(%rip), %rdx
+loop:
+	testq	%rdx, %rdx
+	jz	done
+
+	/* get addresses from the pbe and copy the page */
+	movq	pbe_address(%rdx), %rsi
+	movq	pbe_orig_address(%rdx), %rdi
+	movq	$512, %rcx
+	rep
+	movsq
+
+	/* progress to the next pbe */
+	movq	pbe_next(%rdx), %rdx
+	jmp	loop
+done:
+	/* go back to the original page tables */
+	movq    $(init_level4_pgt - __START_KERNEL_map), %rax
+	addq    phys_base(%rip), %rax
+	movq    %rax, %cr3
+
+	/* Flush TLB, including "global" things (vmalloc) */
+	movq	mmu_cr4_features(%rip), %rax
+	movq	%rax, %rdx
+	andq	$~(1<<7), %rdx;  # PGE
+	movq	%rdx, %cr4;  # turn off PGE
+	movq	%cr3, %rcx;  # flush TLB
+	movq	%rcx, %cr3
+	movq	%rax, %cr4;  # turn PGE back on
+
+	movl	$24, %eax
+	movl	%eax, %ds
+
+	movq saved_context_esp(%rip), %rsp
+	movq saved_context_ebp(%rip), %rbp
+	/* Don't restore %rax, it must be 0 anyway */
+	movq saved_context_ebx(%rip), %rbx
+	movq saved_context_ecx(%rip), %rcx
+	movq saved_context_edx(%rip), %rdx
+	movq saved_context_esi(%rip), %rsi
+	movq saved_context_edi(%rip), %rdi
+	movq saved_context_r08(%rip), %r8
+	movq saved_context_r09(%rip), %r9
+	movq saved_context_r10(%rip), %r10
+	movq saved_context_r11(%rip), %r11
+	movq saved_context_r12(%rip), %r12
+	movq saved_context_r13(%rip), %r13
+	movq saved_context_r14(%rip), %r14
+	movq saved_context_r15(%rip), %r15
+	pushq saved_context_eflags(%rip) ; popfq
+
+	xorq	%rax, %rax
+
+	ret

arch/x86/kernel/sys_x86_64.c

@@ -0,0 +1,159 @@
+/*
+ * linux/arch/x86_64/kernel/sys_x86_64.c
+ */
+
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/syscalls.h>
+#include <linux/mm.h>
+#include <linux/fs.h>
+#include <linux/smp.h>
+#include <linux/sem.h>
+#include <linux/msg.h>
+#include <linux/shm.h>
+#include <linux/stat.h>
+#include <linux/mman.h>
+#include <linux/file.h>
+#include <linux/utsname.h>
+#include <linux/personality.h>
+
+#include <asm/uaccess.h>
+#include <asm/ia32.h>
+
+/*
+ * sys_pipe() is the normal C calling standard for creating
+ * a pipe. It's not the way Unix traditionally does this, though.
+ */
+asmlinkage long sys_pipe(int __user *fildes)
+{
+	int fd[2];
+	int error;
+
+	error = do_pipe(fd);
+	if (!error) {
+		if (copy_to_user(fildes, fd, 2*sizeof(int)))
+			error = -EFAULT;
+	}
+	return error;
+}
+
+asmlinkage long sys_mmap(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags,
+	unsigned long fd, unsigned long off)
+{
+	long error;
+	struct file * file;
+
+	error = -EINVAL;
+	if (off & ~PAGE_MASK)
+		goto out;
+
+	error = -EBADF;
+	file = NULL;
+	flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);
+	if (!(flags & MAP_ANONYMOUS)) {
+		file = fget(fd);
+		if (!file)
+			goto out;
+	}
+	down_write(&current->mm->mmap_sem);
+	error = do_mmap_pgoff(file, addr, len, prot, flags, off >> PAGE_SHIFT);
+	up_write(&current->mm->mmap_sem);
+
+	if (file)
+		fput(file);
+out:
+	return error;
+}
+
+static void find_start_end(unsigned long flags, unsigned long *begin,
+			   unsigned long *end)
+{
+	if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) {
+		/* This is usually used needed to map code in small
+		   model, so it needs to be in the first 31bit. Limit
+		   it to that.  This means we need to move the
+		   unmapped base down for this case. This can give
+		   conflicts with the heap, but we assume that glibc
+		   malloc knows how to fall back to mmap. Give it 1GB
+		   of playground for now. -AK */ 
+		*begin = 0x40000000; 
+		*end = 0x80000000;		
+	} else {
+		*begin = TASK_UNMAPPED_BASE;
+		*end = TASK_SIZE; 
+	}
+} 
+
+unsigned long
+arch_get_unmapped_area(struct file *filp, unsigned long addr,
+		unsigned long len, unsigned long pgoff, unsigned long flags)
+{
+	struct mm_struct *mm = current->mm;
+	struct vm_area_struct *vma;
+	unsigned long start_addr;
+	unsigned long begin, end;
+	
+	if (flags & MAP_FIXED)
+		return addr;
+
+	find_start_end(flags, &begin, &end); 
+
+	if (len > end)
+		return -ENOMEM;
+
+	if (addr) {
+		addr = PAGE_ALIGN(addr);
+		vma = find_vma(mm, addr);
+		if (end - len >= addr &&
+		    (!vma || addr + len <= vma->vm_start))
+			return addr;
+	}
+	if (((flags & MAP_32BIT) || test_thread_flag(TIF_IA32))
+	    && len <= mm->cached_hole_size) {
+	        mm->cached_hole_size = 0;
+		mm->free_area_cache = begin;
+	}
+	addr = mm->free_area_cache;
+	if (addr < begin) 
+		addr = begin; 
+	start_addr = addr;
+
+full_search:
+	for (vma = find_vma(mm, addr); ; vma = vma->vm_next) {
+		/* At this point:  (!vma || addr < vma->vm_end). */
+		if (end - len < addr) {
+			/*
+			 * Start a new search - just in case we missed
+			 * some holes.
+			 */
+			if (start_addr != begin) {
+				start_addr = addr = begin;
+				mm->cached_hole_size = 0;
+				goto full_search;
+			}
+			return -ENOMEM;
+		}
+		if (!vma || addr + len <= vma->vm_start) {
+			/*
+			 * Remember the place where we stopped the search:
+			 */
+			mm->free_area_cache = addr + len;
+			return addr;
+		}
+		if (addr + mm->cached_hole_size < vma->vm_start)
+		        mm->cached_hole_size = vma->vm_start - addr;
+
+		addr = vma->vm_end;
+	}
+}
+
+asmlinkage long sys_uname(struct new_utsname __user * name)
+{
+	int err;
+	down_read(&uts_sem);
+	err = copy_to_user(name, utsname(), sizeof (*name));
+	up_read(&uts_sem);
+	if (personality(current->personality) == PER_LINUX32) 
+		err |= copy_to_user(&name->machine, "i686", 5); 		
+	return err ? -EFAULT : 0;
+}

arch/x86/kernel/syscall_64.c

@@ -0,0 +1,26 @@
+/* System call table for x86-64. */ 
+
+#include <linux/linkage.h>
+#include <linux/sys.h>
+#include <linux/cache.h>
+#include <asm/asm-offsets.h>
+
+#define __NO_STUBS
+
+#define __SYSCALL(nr, sym) extern asmlinkage void sym(void) ; 
+#undef _ASM_X86_64_UNISTD_H_
+#include <asm-x86_64/unistd.h>
+
+#undef __SYSCALL
+#define __SYSCALL(nr, sym) [ nr ] = sym, 
+#undef _ASM_X86_64_UNISTD_H_
+
+typedef void (*sys_call_ptr_t)(void); 
+
+extern void sys_ni_syscall(void);
+
+const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = {
+	/* Smells like a like a compiler bug -- it doesn't work when the & below is removed. */ 
+	[0 ... __NR_syscall_max] = &sys_ni_syscall,
+#include <asm-x86_64/unistd.h>
+};

arch/x86/kernel/tce_64.c

@@ -0,0 +1,189 @@
+/*
+ * This file manages the translation entries for the IBM Calgary IOMMU.
+ *
+ * Derived from arch/powerpc/platforms/pseries/iommu.c
+ *
+ * Copyright (C) IBM Corporation, 2006
+ *
+ * Author: Jon Mason <jdmason@us.ibm.com>
+ * Author: Muli Ben-Yehuda <muli@il.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ */
+
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/mm.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <linux/pci.h>
+#include <linux/dma-mapping.h>
+#include <linux/bootmem.h>
+#include <asm/tce.h>
+#include <asm/calgary.h>
+#include <asm/proto.h>
+
+/* flush a tce at 'tceaddr' to main memory */
+static inline void flush_tce(void* tceaddr)
+{
+	/* a single tce can't cross a cache line */
+	if (cpu_has_clflush)
+		asm volatile("clflush (%0)" :: "r" (tceaddr));
+	else
+		asm volatile("wbinvd":::"memory");
+}
+
+void tce_build(struct iommu_table *tbl, unsigned long index,
+	unsigned int npages, unsigned long uaddr, int direction)
+{
+	u64* tp;
+	u64 t;
+	u64 rpn;
+
+	t = (1 << TCE_READ_SHIFT);
+	if (direction != DMA_TO_DEVICE)
+		t |= (1 << TCE_WRITE_SHIFT);
+
+	tp = ((u64*)tbl->it_base) + index;
+
+	while (npages--) {
+		rpn = (virt_to_bus((void*)uaddr)) >> PAGE_SHIFT;
+		t &= ~TCE_RPN_MASK;
+		t |= (rpn << TCE_RPN_SHIFT);
+
+		*tp = cpu_to_be64(t);
+		flush_tce(tp);
+
+		uaddr += PAGE_SIZE;
+		tp++;
+	}
+}
+
+void tce_free(struct iommu_table *tbl, long index, unsigned int npages)
+{
+	u64* tp;
+
+	tp  = ((u64*)tbl->it_base) + index;
+
+	while (npages--) {
+		*tp = cpu_to_be64(0);
+		flush_tce(tp);
+		tp++;
+	}
+}
+
+static inline unsigned int table_size_to_number_of_entries(unsigned char size)
+{
+	/*
+	 * size is the order of the table, 0-7
+	 * smallest table is 8K entries, so shift result by 13 to
+	 * multiply by 8K
+	 */
+	return (1 << size) << 13;
+}
+
+static int tce_table_setparms(struct pci_dev *dev, struct iommu_table *tbl)
+{
+	unsigned int bitmapsz;
+	unsigned long bmppages;
+	int ret;
+
+	tbl->it_busno = dev->bus->number;
+
+	/* set the tce table size - measured in entries */
+	tbl->it_size = table_size_to_number_of_entries(specified_table_size);
+
+	/*
+	 * number of bytes needed for the bitmap size in number of
+	 * entries; we need one bit per entry
+	 */
+	bitmapsz = tbl->it_size / BITS_PER_BYTE;
+	bmppages = __get_free_pages(GFP_KERNEL, get_order(bitmapsz));
+	if (!bmppages) {
+		printk(KERN_ERR "Calgary: cannot allocate bitmap\n");
+		ret = -ENOMEM;
+		goto done;
+	}
+
+	tbl->it_map = (unsigned long*)bmppages;
+
+	memset(tbl->it_map, 0, bitmapsz);
+
+	tbl->it_hint = 0;
+
+	spin_lock_init(&tbl->it_lock);
+
+	return 0;
+
+done:
+	return ret;
+}
+
+int __init build_tce_table(struct pci_dev *dev, void __iomem *bbar)
+{
+	struct iommu_table *tbl;
+	int ret;
+
+	if (pci_iommu(dev->bus)) {
+		printk(KERN_ERR "Calgary: dev %p has sysdata->iommu %p\n",
+		       dev, pci_iommu(dev->bus));
+		BUG();
+	}
+
+	tbl = kzalloc(sizeof(struct iommu_table), GFP_KERNEL);
+	if (!tbl) {
+		printk(KERN_ERR "Calgary: error allocating iommu_table\n");
+		ret = -ENOMEM;
+		goto done;
+	}
+
+	ret = tce_table_setparms(dev, tbl);
+	if (ret)
+		goto free_tbl;
+
+	tbl->bbar = bbar;
+
+	set_pci_iommu(dev->bus, tbl);
+
+	return 0;
+
+free_tbl:
+	kfree(tbl);
+done:
+	return ret;
+}
+
+void * __init alloc_tce_table(void)
+{
+	unsigned int size;
+
+	size = table_size_to_number_of_entries(specified_table_size);
+	size *= TCE_ENTRY_SIZE;
+
+	return __alloc_bootmem_low(size, size, 0);
+}
+
+void __init free_tce_table(void *tbl)
+{
+	unsigned int size;
+
+	if (!tbl)
+		return;
+
+	size = table_size_to_number_of_entries(specified_table_size);
+	size *= TCE_ENTRY_SIZE;
+
+	free_bootmem(__pa(tbl), size);
+}

arch/x86/kernel/time_64.c

@@ -0,0 +1,447 @@
+/*
+ *  linux/arch/x86-64/kernel/time.c
+ *
+ *  "High Precision Event Timer" based timekeeping.
+ *
+ *  Copyright (c) 1991,1992,1995  Linus Torvalds
+ *  Copyright (c) 1994  Alan Modra
+ *  Copyright (c) 1995  Markus Kuhn
+ *  Copyright (c) 1996  Ingo Molnar
+ *  Copyright (c) 1998  Andrea Arcangeli
+ *  Copyright (c) 2002,2006  Vojtech Pavlik
+ *  Copyright (c) 2003  Andi Kleen
+ *  RTC support code taken from arch/i386/kernel/timers/time_hpet.c
+ */
+
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/interrupt.h>
+#include <linux/init.h>
+#include <linux/mc146818rtc.h>
+#include <linux/time.h>
+#include <linux/ioport.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/sysdev.h>
+#include <linux/bcd.h>
+#include <linux/notifier.h>
+#include <linux/cpu.h>
+#include <linux/kallsyms.h>
+#include <linux/acpi.h>
+#ifdef CONFIG_ACPI
+#include <acpi/achware.h>	/* for PM timer frequency */
+#include <acpi/acpi_bus.h>
+#endif
+#include <asm/8253pit.h>
+#include <asm/i8253.h>
+#include <asm/pgtable.h>
+#include <asm/vsyscall.h>
+#include <asm/timex.h>
+#include <asm/proto.h>
+#include <asm/hpet.h>
+#include <asm/sections.h>
+#include <linux/hpet.h>
+#include <asm/apic.h>
+#include <asm/hpet.h>
+#include <asm/mpspec.h>
+#include <asm/nmi.h>
+#include <asm/vgtod.h>
+
+static char *timename = NULL;
+
+DEFINE_SPINLOCK(rtc_lock);
+EXPORT_SYMBOL(rtc_lock);
+DEFINE_SPINLOCK(i8253_lock);
+EXPORT_SYMBOL(i8253_lock);
+
+volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
+
+unsigned long profile_pc(struct pt_regs *regs)
+{
+	unsigned long pc = instruction_pointer(regs);
+
+	/* Assume the lock function has either no stack frame or a copy
+	   of eflags from PUSHF
+	   Eflags always has bits 22 and up cleared unlike kernel addresses. */
+	if (!user_mode(regs) && in_lock_functions(pc)) {
+		unsigned long *sp = (unsigned long *)regs->rsp;
+		if (sp[0] >> 22)
+			return sp[0];
+		if (sp[1] >> 22)
+			return sp[1];
+	}
+	return pc;
+}
+EXPORT_SYMBOL(profile_pc);
+
+/*
+ * In order to set the CMOS clock precisely, set_rtc_mmss has to be called 500
+ * ms after the second nowtime has started, because when nowtime is written
+ * into the registers of the CMOS clock, it will jump to the next second
+ * precisely 500 ms later. Check the Motorola MC146818A or Dallas DS12887 data
+ * sheet for details.
+ */
+
+static int set_rtc_mmss(unsigned long nowtime)
+{
+	int retval = 0;
+	int real_seconds, real_minutes, cmos_minutes;
+	unsigned char control, freq_select;
+
+/*
+ * IRQs are disabled when we're called from the timer interrupt,
+ * no need for spin_lock_irqsave()
+ */
+
+	spin_lock(&rtc_lock);
+
+/*
+ * Tell the clock it's being set and stop it.
+ */
+
+	control = CMOS_READ(RTC_CONTROL);
+	CMOS_WRITE(control | RTC_SET, RTC_CONTROL);
+
+	freq_select = CMOS_READ(RTC_FREQ_SELECT);
+	CMOS_WRITE(freq_select | RTC_DIV_RESET2, RTC_FREQ_SELECT);
+
+	cmos_minutes = CMOS_READ(RTC_MINUTES);
+		BCD_TO_BIN(cmos_minutes);
+
+/*
+ * since we're only adjusting minutes and seconds, don't interfere with hour
+ * overflow. This avoids messing with unknown time zones but requires your RTC
+ * not to be off by more than 15 minutes. Since we're calling it only when
+ * our clock is externally synchronized using NTP, this shouldn't be a problem.
+ */
+
+	real_seconds = nowtime % 60;
+	real_minutes = nowtime / 60;
+	if (((abs(real_minutes - cmos_minutes) + 15) / 30) & 1)
+		real_minutes += 30;		/* correct for half hour time zone */
+	real_minutes %= 60;
+
+	if (abs(real_minutes - cmos_minutes) >= 30) {
+		printk(KERN_WARNING "time.c: can't update CMOS clock "
+		       "from %d to %d\n", cmos_minutes, real_minutes);
+		retval = -1;
+	} else {
+		BIN_TO_BCD(real_seconds);
+		BIN_TO_BCD(real_minutes);
+		CMOS_WRITE(real_seconds, RTC_SECONDS);
+		CMOS_WRITE(real_minutes, RTC_MINUTES);
+	}
+
+/*
+ * The following flags have to be released exactly in this order, otherwise the
+ * DS12887 (popular MC146818A clone with integrated battery and quartz) will
+ * not reset the oscillator and will not update precisely 500 ms later. You
+ * won't find this mentioned in the Dallas Semiconductor data sheets, but who
+ * believes data sheets anyway ... -- Markus Kuhn
+ */
+
+	CMOS_WRITE(control, RTC_CONTROL);
+	CMOS_WRITE(freq_select, RTC_FREQ_SELECT);
+
+	spin_unlock(&rtc_lock);
+
+	return retval;
+}
+
+int update_persistent_clock(struct timespec now)
+{
+	return set_rtc_mmss(now.tv_sec);
+}
+
+void main_timer_handler(void)
+{
+/*
+ * Here we are in the timer irq handler. We have irqs locally disabled (so we
+ * don't need spin_lock_irqsave()) but we don't know if the timer_bh is running
+ * on the other CPU, so we need a lock. We also need to lock the vsyscall
+ * variables, because both do_timer() and us change them -arca+vojtech
+ */
+
+	write_seqlock(&xtime_lock);
+
+/*
+ * Do the timer stuff.
+ */
+
+	do_timer(1);
+#ifndef CONFIG_SMP
+	update_process_times(user_mode(get_irq_regs()));
+#endif
+
+/*
+ * In the SMP case we use the local APIC timer interrupt to do the profiling,
+ * except when we simulate SMP mode on a uniprocessor system, in that case we
+ * have to call the local interrupt handler.
+ */
+
+	if (!using_apic_timer)
+		smp_local_timer_interrupt();
+
+	write_sequnlock(&xtime_lock);
+}
+
+static irqreturn_t timer_interrupt(int irq, void *dev_id)
+{
+	if (apic_runs_main_timer > 1)
+		return IRQ_HANDLED;
+	main_timer_handler();
+	if (using_apic_timer)
+		smp_send_timer_broadcast_ipi();
+	return IRQ_HANDLED;
+}
+
+unsigned long read_persistent_clock(void)
+{
+	unsigned int year, mon, day, hour, min, sec;
+	unsigned long flags;
+	unsigned century = 0;
+
+	spin_lock_irqsave(&rtc_lock, flags);
+
+	do {
+		sec = CMOS_READ(RTC_SECONDS);
+		min = CMOS_READ(RTC_MINUTES);
+		hour = CMOS_READ(RTC_HOURS);
+		day = CMOS_READ(RTC_DAY_OF_MONTH);
+		mon = CMOS_READ(RTC_MONTH);
+		year = CMOS_READ(RTC_YEAR);
+#ifdef CONFIG_ACPI
+		if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID &&
+					acpi_gbl_FADT.century)
+			century = CMOS_READ(acpi_gbl_FADT.century);
+#endif
+	} while (sec != CMOS_READ(RTC_SECONDS));
+
+	spin_unlock_irqrestore(&rtc_lock, flags);
+
+	/*
+	 * We know that x86-64 always uses BCD format, no need to check the
+	 * config register.
+	 */
+
+	BCD_TO_BIN(sec);
+	BCD_TO_BIN(min);
+	BCD_TO_BIN(hour);
+	BCD_TO_BIN(day);
+	BCD_TO_BIN(mon);
+	BCD_TO_BIN(year);
+
+	if (century) {
+		BCD_TO_BIN(century);
+		year += century * 100;
+		printk(KERN_INFO "Extended CMOS year: %d\n", century * 100);
+	} else {
+		/*
+		 * x86-64 systems only exists since 2002.
+		 * This will work up to Dec 31, 2100
+		 */
+		year += 2000;
+	}
+
+	return mktime(year, mon, day, hour, min, sec);
+}
+
+/* calibrate_cpu is used on systems with fixed rate TSCs to determine
+ * processor frequency */
+#define TICK_COUNT 100000000
+static unsigned int __init tsc_calibrate_cpu_khz(void)
+{
+	int tsc_start, tsc_now;
+	int i, no_ctr_free;
+	unsigned long evntsel3 = 0, pmc3 = 0, pmc_now = 0;
+	unsigned long flags;
+
+	for (i = 0; i < 4; i++)
+		if (avail_to_resrv_perfctr_nmi_bit(i))
+			break;
+	no_ctr_free = (i == 4);
+	if (no_ctr_free) {
+		i = 3;
+		rdmsrl(MSR_K7_EVNTSEL3, evntsel3);
+		wrmsrl(MSR_K7_EVNTSEL3, 0);
+		rdmsrl(MSR_K7_PERFCTR3, pmc3);
+	} else {
+		reserve_perfctr_nmi(MSR_K7_PERFCTR0 + i);
+		reserve_evntsel_nmi(MSR_K7_EVNTSEL0 + i);
+	}
+	local_irq_save(flags);
+	/* start meauring cycles, incrementing from 0 */
+	wrmsrl(MSR_K7_PERFCTR0 + i, 0);
+	wrmsrl(MSR_K7_EVNTSEL0 + i, 1 << 22 | 3 << 16 | 0x76);
+	rdtscl(tsc_start);
+	do {
+		rdmsrl(MSR_K7_PERFCTR0 + i, pmc_now);
+		tsc_now = get_cycles_sync();
+	} while ((tsc_now - tsc_start) < TICK_COUNT);
+
+	local_irq_restore(flags);
+	if (no_ctr_free) {
+		wrmsrl(MSR_K7_EVNTSEL3, 0);
+		wrmsrl(MSR_K7_PERFCTR3, pmc3);
+		wrmsrl(MSR_K7_EVNTSEL3, evntsel3);
+	} else {
+		release_perfctr_nmi(MSR_K7_PERFCTR0 + i);
+		release_evntsel_nmi(MSR_K7_EVNTSEL0 + i);
+	}
+
+	return pmc_now * tsc_khz / (tsc_now - tsc_start);
+}
+
+/*
+ * pit_calibrate_tsc() uses the speaker output (channel 2) of
+ * the PIT. This is better than using the timer interrupt output,
+ * because we can read the value of the speaker with just one inb(),
+ * where we need three i/o operations for the interrupt channel.
+ * We count how many ticks the TSC does in 50 ms.
+ */
+
+static unsigned int __init pit_calibrate_tsc(void)
+{
+	unsigned long start, end;
+	unsigned long flags;
+
+	spin_lock_irqsave(&i8253_lock, flags);
+
+	outb((inb(0x61) & ~0x02) | 0x01, 0x61);
+
+	outb(0xb0, 0x43);
+	outb((PIT_TICK_RATE / (1000 / 50)) & 0xff, 0x42);
+	outb((PIT_TICK_RATE / (1000 / 50)) >> 8, 0x42);
+	start = get_cycles_sync();
+	while ((inb(0x61) & 0x20) == 0);
+	end = get_cycles_sync();
+
+	spin_unlock_irqrestore(&i8253_lock, flags);
+
+	return (end - start) / 50;
+}
+
+#define PIT_MODE 0x43
+#define PIT_CH0  0x40
+
+static void __pit_init(int val, u8 mode)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&i8253_lock, flags);
+	outb_p(mode, PIT_MODE);
+	outb_p(val & 0xff, PIT_CH0);	/* LSB */
+	outb_p(val >> 8, PIT_CH0);	/* MSB */
+	spin_unlock_irqrestore(&i8253_lock, flags);
+}
+
+void __init pit_init(void)
+{
+	__pit_init(LATCH, 0x34); /* binary, mode 2, LSB/MSB, ch 0 */
+}
+
+void pit_stop_interrupt(void)
+{
+	__pit_init(0, 0x30); /* mode 0 */
+}
+
+void stop_timer_interrupt(void)
+{
+	char *name;
+	if (hpet_address) {
+		name = "HPET";
+		hpet_timer_stop_set_go(0);
+	} else {
+		name = "PIT";
+		pit_stop_interrupt();
+	}
+	printk(KERN_INFO "timer: %s interrupt stopped.\n", name);
+}
+
+static struct irqaction irq0 = {
+	.handler	= timer_interrupt,
+	.flags		= IRQF_DISABLED | IRQF_IRQPOLL,
+	.mask		= CPU_MASK_NONE,
+	.name		= "timer"
+};
+
+void __init time_init(void)
+{
+	if (nohpet)
+		hpet_address = 0;
+
+	if (hpet_arch_init())
+		hpet_address = 0;
+
+	if (hpet_use_timer) {
+		/* set tick_nsec to use the proper rate for HPET */
+		tick_nsec = TICK_NSEC_HPET;
+		tsc_khz = hpet_calibrate_tsc();
+		timename = "HPET";
+	} else {
+		pit_init();
+		tsc_khz = pit_calibrate_tsc();
+		timename = "PIT";
+	}
+
+	cpu_khz = tsc_khz;
+	if (cpu_has(&boot_cpu_data, X86_FEATURE_CONSTANT_TSC) &&
+		boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
+		boot_cpu_data.x86 == 16)
+		cpu_khz = tsc_calibrate_cpu_khz();
+
+	if (unsynchronized_tsc())
+		mark_tsc_unstable("TSCs unsynchronized");
+
+	if (cpu_has(&boot_cpu_data, X86_FEATURE_RDTSCP))
+		vgetcpu_mode = VGETCPU_RDTSCP;
+	else
+		vgetcpu_mode = VGETCPU_LSL;
+
+	set_cyc2ns_scale(tsc_khz);
+	printk(KERN_INFO "time.c: Detected %d.%03d MHz processor.\n",
+		cpu_khz / 1000, cpu_khz % 1000);
+	init_tsc_clocksource();
+
+	setup_irq(0, &irq0);
+}
+
+/*
+ * sysfs support for the timer.
+ */
+
+static int timer_suspend(struct sys_device *dev, pm_message_t state)
+{
+	return 0;
+}
+
+static int timer_resume(struct sys_device *dev)
+{
+	if (hpet_address)
+		hpet_reenable();
+	else
+		i8254_timer_resume();
+	return 0;
+}
+
+static struct sysdev_class timer_sysclass = {
+	.resume = timer_resume,
+	.suspend = timer_suspend,
+	set_kset_name("timer"),
+};
+
+/* XXX this sysfs stuff should probably go elsewhere later -john */
+static struct sys_device device_timer = {
+	.id	= 0,
+	.cls	= &timer_sysclass,
+};
+
+static int time_init_device(void)
+{
+	int error = sysdev_class_register(&timer_sysclass);
+	if (!error)
+		error = sysdev_register(&device_timer);
+	return error;
+}
+
+device_initcall(time_init_device);

arch/x86/kernel/trampoline_64.S

@@ -0,0 +1,166 @@
+/*
+ *
+ *	Trampoline.S	Derived from Setup.S by Linus Torvalds
+ *
+ *	4 Jan 1997 Michael Chastain: changed to gnu as.
+ *	15 Sept 2005 Eric Biederman: 64bit PIC support
+ *
+ *	Entry: CS:IP point to the start of our code, we are 
+ *	in real mode with no stack, but the rest of the 
+ *	trampoline page to make our stack and everything else
+ *	is a mystery.
+ *
+ *	In fact we don't actually need a stack so we don't
+ *	set one up.
+ *
+ *	On entry to trampoline_data, the processor is in real mode
+ *	with 16-bit addressing and 16-bit data.  CS has some value
+ *	and IP is zero.  Thus, data addresses need to be absolute
+ *	(no relocation) and are taken with regard to r_base.
+ *
+ *	With the addition of trampoline_level4_pgt this code can
+ *	now enter a 64bit kernel that lives at arbitrary 64bit
+ *	physical addresses.
+ *
+ *	If you work on this file, check the object module with objdump
+ *	--full-contents --reloc to make sure there are no relocation
+ *	entries.
+ */
+
+#include <linux/linkage.h>
+#include <asm/pgtable.h>
+#include <asm/page.h>
+#include <asm/msr.h>
+#include <asm/segment.h>
+
+.data
+
+.code16
+
+ENTRY(trampoline_data)
+r_base = .
+	cli			# We should be safe anyway
+	wbinvd	
+	mov	%cs, %ax	# Code and data in the same place
+	mov	%ax, %ds
+	mov	%ax, %es
+	mov	%ax, %ss
+
+
+	movl	$0xA5A5A5A5, trampoline_data - r_base
+				# write marker for master knows we're running
+
+					# Setup stack
+	movw	$(trampoline_stack_end - r_base), %sp
+
+	call	verify_cpu		# Verify the cpu supports long mode
+	testl   %eax, %eax		# Check for return code
+	jnz	no_longmode
+
+	mov	%cs, %ax
+	movzx	%ax, %esi		# Find the 32bit trampoline location
+	shll	$4, %esi
+
+					# Fixup the vectors
+	addl	%esi, startup_32_vector - r_base
+	addl	%esi, startup_64_vector - r_base
+	addl	%esi, tgdt + 2 - r_base	# Fixup the gdt pointer
+
+	/*
+	 * GDT tables in non default location kernel can be beyond 16MB and
+	 * lgdt will not be able to load the address as in real mode default
+	 * operand size is 16bit. Use lgdtl instead to force operand size
+	 * to 32 bit.
+	 */
+
+	lidtl	tidt - r_base	# load idt with 0, 0
+	lgdtl	tgdt - r_base	# load gdt with whatever is appropriate
+
+	xor	%ax, %ax
+	inc	%ax		# protected mode (PE) bit
+	lmsw	%ax		# into protected mode
+
+	# flush prefetch and jump to startup_32
+	ljmpl	*(startup_32_vector - r_base)
+
+	.code32
+	.balign 4
+startup_32:
+	movl	$__KERNEL_DS, %eax	# Initialize the %ds segment register
+	movl	%eax, %ds
+
+	xorl	%eax, %eax
+	btsl	$5, %eax		# Enable PAE mode
+	movl	%eax, %cr4
+
+					# Setup trampoline 4 level pagetables
+	leal	(trampoline_level4_pgt - r_base)(%esi), %eax
+	movl	%eax, %cr3
+
+	movl	$MSR_EFER, %ecx
+	movl	$(1 << _EFER_LME), %eax	# Enable Long Mode
+	xorl	%edx, %edx
+	wrmsr
+
+	xorl	%eax, %eax
+	btsl	$31, %eax		# Enable paging and in turn activate Long Mode
+	btsl	$0, %eax		# Enable protected mode
+	movl	%eax, %cr0
+
+	/*
+	 * At this point we're in long mode but in 32bit compatibility mode
+	 * with EFER.LME = 1, CS.L = 0, CS.D = 1 (and in turn
+	 * EFER.LMA = 1). Now we want to jump in 64bit mode, to do that we use
+	 * the new gdt/idt that has __KERNEL_CS with CS.L = 1.
+	 */
+	ljmp	*(startup_64_vector - r_base)(%esi)
+
+	.code64
+	.balign 4
+startup_64:
+	# Now jump into the kernel using virtual addresses
+	movq	$secondary_startup_64, %rax
+	jmp	*%rax
+
+	.code16
+no_longmode:
+	hlt
+	jmp no_longmode
+#include "verify_cpu_64.S"
+
+	# Careful these need to be in the same 64K segment as the above;
+tidt:
+	.word	0			# idt limit = 0
+	.word	0, 0			# idt base = 0L
+
+	# Duplicate the global descriptor table
+	# so the kernel can live anywhere
+	.balign 4
+tgdt:
+	.short	tgdt_end - tgdt		# gdt limit
+	.long	tgdt - r_base
+	.short 0
+	.quad	0x00cf9b000000ffff	# __KERNEL32_CS
+	.quad	0x00af9b000000ffff	# __KERNEL_CS
+	.quad	0x00cf93000000ffff	# __KERNEL_DS
+tgdt_end:
+
+	.balign 4
+startup_32_vector:
+	.long	startup_32 - r_base
+	.word	__KERNEL32_CS, 0
+
+	.balign 4
+startup_64_vector:
+	.long	startup_64 - r_base
+	.word	__KERNEL_CS, 0
+
+trampoline_stack:
+	.org 0x1000
+trampoline_stack_end:
+ENTRY(trampoline_level4_pgt)
+	.quad	level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE
+	.fill	510,8,0
+	.quad	level3_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE
+
+ENTRY(trampoline_end)

arch/x86/kernel/tsc_64.c

@@ -0,0 +1,207 @@
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/interrupt.h>
+#include <linux/init.h>
+#include <linux/clocksource.h>
+#include <linux/time.h>
+#include <linux/acpi.h>
+#include <linux/cpufreq.h>
+
+#include <asm/timex.h>
+
+static int notsc __initdata = 0;
+
+unsigned int cpu_khz;		/* TSC clocks / usec, not used here */
+EXPORT_SYMBOL(cpu_khz);
+unsigned int tsc_khz;
+EXPORT_SYMBOL(tsc_khz);
+
+static unsigned int cyc2ns_scale __read_mostly;
+
+void set_cyc2ns_scale(unsigned long khz)
+{
+	cyc2ns_scale = (NSEC_PER_MSEC << NS_SCALE) / khz;
+}
+
+static unsigned long long cycles_2_ns(unsigned long long cyc)
+{
+	return (cyc * cyc2ns_scale) >> NS_SCALE;
+}
+
+unsigned long long sched_clock(void)
+{
+	unsigned long a = 0;
+
+	/* Could do CPU core sync here. Opteron can execute rdtsc speculatively,
+	 * which means it is not completely exact and may not be monotonous
+	 * between CPUs. But the errors should be too small to matter for
+	 * scheduling purposes.
+	 */
+
+	rdtscll(a);
+	return cycles_2_ns(a);
+}
+
+static int tsc_unstable;
+
+inline int check_tsc_unstable(void)
+{
+	return tsc_unstable;
+}
+#ifdef CONFIG_CPU_FREQ
+
+/* Frequency scaling support. Adjust the TSC based timer when the cpu frequency
+ * changes.
+ *
+ * RED-PEN: On SMP we assume all CPUs run with the same frequency.  It's
+ * not that important because current Opteron setups do not support
+ * scaling on SMP anyroads.
+ *
+ * Should fix up last_tsc too. Currently gettimeofday in the
+ * first tick after the change will be slightly wrong.
+ */
+
+static unsigned int  ref_freq;
+static unsigned long loops_per_jiffy_ref;
+static unsigned long tsc_khz_ref;
+
+static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
+				 void *data)
+{
+	struct cpufreq_freqs *freq = data;
+	unsigned long *lpj, dummy;
+
+	if (cpu_has(&cpu_data[freq->cpu], X86_FEATURE_CONSTANT_TSC))
+		return 0;
+
+	lpj = &dummy;
+	if (!(freq->flags & CPUFREQ_CONST_LOOPS))
+#ifdef CONFIG_SMP
+		lpj = &cpu_data[freq->cpu].loops_per_jiffy;
+#else
+		lpj = &boot_cpu_data.loops_per_jiffy;
+#endif
+
+	if (!ref_freq) {
+		ref_freq = freq->old;
+		loops_per_jiffy_ref = *lpj;
+		tsc_khz_ref = tsc_khz;
+	}
+	if ((val == CPUFREQ_PRECHANGE  && freq->old < freq->new) ||
+		(val == CPUFREQ_POSTCHANGE && freq->old > freq->new) ||
+		(val == CPUFREQ_RESUMECHANGE)) {
+		*lpj =
+		cpufreq_scale(loops_per_jiffy_ref, ref_freq, freq->new);
+
+		tsc_khz = cpufreq_scale(tsc_khz_ref, ref_freq, freq->new);
+		if (!(freq->flags & CPUFREQ_CONST_LOOPS))
+			mark_tsc_unstable("cpufreq changes");
+	}
+
+	set_cyc2ns_scale(tsc_khz_ref);
+
+	return 0;
+}
+
+static struct notifier_block time_cpufreq_notifier_block = {
+	.notifier_call  = time_cpufreq_notifier
+};
+
+static int __init cpufreq_tsc(void)
+{
+	cpufreq_register_notifier(&time_cpufreq_notifier_block,
+				  CPUFREQ_TRANSITION_NOTIFIER);
+	return 0;
+}
+
+core_initcall(cpufreq_tsc);
+
+#endif
+
+/*
+ * Make an educated guess if the TSC is trustworthy and synchronized
+ * over all CPUs.
+ */
+__cpuinit int unsynchronized_tsc(void)
+{
+	if (tsc_unstable)
+		return 1;
+
+#ifdef CONFIG_SMP
+	if (apic_is_clustered_box())
+		return 1;
+#endif
+	/* Most intel systems have synchronized TSCs except for
+	   multi node systems */
+	if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) {
+#ifdef CONFIG_ACPI
+		/* But TSC doesn't tick in C3 so don't use it there */
+		if (acpi_gbl_FADT.header.length > 0 &&
+		    acpi_gbl_FADT.C3latency < 1000)
+			return 1;
+#endif
+		return 0;
+	}
+
+	/* Assume multi socket systems are not synchronized */
+	return num_present_cpus() > 1;
+}
+
+int __init notsc_setup(char *s)
+{
+	notsc = 1;
+	return 1;
+}
+
+__setup("notsc", notsc_setup);
+
+
+/* clock source code: */
+static cycle_t read_tsc(void)
+{
+	cycle_t ret = (cycle_t)get_cycles_sync();
+	return ret;
+}
+
+static cycle_t __vsyscall_fn vread_tsc(void)
+{
+	cycle_t ret = (cycle_t)get_cycles_sync();
+	return ret;
+}
+
+static struct clocksource clocksource_tsc = {
+	.name			= "tsc",
+	.rating			= 300,
+	.read			= read_tsc,
+	.mask			= CLOCKSOURCE_MASK(64),
+	.shift			= 22,
+	.flags			= CLOCK_SOURCE_IS_CONTINUOUS |
+				  CLOCK_SOURCE_MUST_VERIFY,
+	.vread			= vread_tsc,
+};
+
+void mark_tsc_unstable(char *reason)
+{
+	if (!tsc_unstable) {
+		tsc_unstable = 1;
+		printk("Marking TSC unstable due to %s\n", reason);
+		/* Change only the rating, when not registered */
+		if (clocksource_tsc.mult)
+			clocksource_change_rating(&clocksource_tsc, 0);
+		else
+			clocksource_tsc.rating = 0;
+	}
+}
+EXPORT_SYMBOL_GPL(mark_tsc_unstable);
+
+void __init init_tsc_clocksource(void)
+{
+	if (!notsc) {
+		clocksource_tsc.mult = clocksource_khz2mult(tsc_khz,
+							clocksource_tsc.shift);
+		if (check_tsc_unstable())
+			clocksource_tsc.rating = 0;
+
+		clocksource_register(&clocksource_tsc);
+	}
+}

arch/x86/kernel/tsc_sync.c

@@ -1 +1,187 @@
-#include "../../x86_64/kernel/tsc_sync.c"
+/*
+ * arch/x86_64/kernel/tsc_sync.c: check TSC synchronization.
+ *
+ * Copyright (C) 2006, Red Hat, Inc., Ingo Molnar
+ *
+ * We check whether all boot CPUs have their TSC's synchronized,
+ * print a warning if not and turn off the TSC clock-source.
+ *
+ * The warp-check is point-to-point between two CPUs, the CPU
+ * initiating the bootup is the 'source CPU', the freshly booting
+ * CPU is the 'target CPU'.
+ *
+ * Only two CPUs may participate - they can enter in any order.
+ * ( The serial nature of the boot logic and the CPU hotplug lock
+ *   protects against more than 2 CPUs entering this code. )
+ */
+#include <linux/spinlock.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/nmi.h>
+#include <asm/tsc.h>
+
+/*
+ * Entry/exit counters that make sure that both CPUs
+ * run the measurement code at once:
+ */
+static __cpuinitdata atomic_t start_count;
+static __cpuinitdata atomic_t stop_count;
+
+/*
+ * We use a raw spinlock in this exceptional case, because
+ * we want to have the fastest, inlined, non-debug version
+ * of a critical section, to be able to prove TSC time-warps:
+ */
+static __cpuinitdata raw_spinlock_t sync_lock = __RAW_SPIN_LOCK_UNLOCKED;
+static __cpuinitdata cycles_t last_tsc;
+static __cpuinitdata cycles_t max_warp;
+static __cpuinitdata int nr_warps;
+
+/*
+ * TSC-warp measurement loop running on both CPUs:
+ */
+static __cpuinit void check_tsc_warp(void)
+{
+	cycles_t start, now, prev, end;
+	int i;
+
+	start = get_cycles_sync();
+	/*
+	 * The measurement runs for 20 msecs:
+	 */
+	end = start + tsc_khz * 20ULL;
+	now = start;
+
+	for (i = 0; ; i++) {
+		/*
+		 * We take the global lock, measure TSC, save the
+		 * previous TSC that was measured (possibly on
+		 * another CPU) and update the previous TSC timestamp.
+		 */
+		__raw_spin_lock(&sync_lock);
+		prev = last_tsc;
+		now = get_cycles_sync();
+		last_tsc = now;
+		__raw_spin_unlock(&sync_lock);
+
+		/*
+		 * Be nice every now and then (and also check whether
+		 * measurement is done [we also insert a 100 million
+		 * loops safety exit, so we dont lock up in case the
+		 * TSC readout is totally broken]):
+		 */
+		if (unlikely(!(i & 7))) {
+			if (now > end || i > 100000000)
+				break;
+			cpu_relax();
+			touch_nmi_watchdog();
+		}
+		/*
+		 * Outside the critical section we can now see whether
+		 * we saw a time-warp of the TSC going backwards:
+		 */
+		if (unlikely(prev > now)) {
+			__raw_spin_lock(&sync_lock);
+			max_warp = max(max_warp, prev - now);
+			nr_warps++;
+			__raw_spin_unlock(&sync_lock);
+		}
+
+	}
+}
+
+/*
+ * Source CPU calls into this - it waits for the freshly booted
+ * target CPU to arrive and then starts the measurement:
+ */
+void __cpuinit check_tsc_sync_source(int cpu)
+{
+	int cpus = 2;
+
+	/*
+	 * No need to check if we already know that the TSC is not
+	 * synchronized:
+	 */
+	if (unsynchronized_tsc())
+		return;
+
+	printk(KERN_INFO "checking TSC synchronization [CPU#%d -> CPU#%d]:",
+			  smp_processor_id(), cpu);
+
+	/*
+	 * Reset it - in case this is a second bootup:
+	 */
+	atomic_set(&stop_count, 0);
+
+	/*
+	 * Wait for the target to arrive:
+	 */
+	while (atomic_read(&start_count) != cpus-1)
+		cpu_relax();
+	/*
+	 * Trigger the target to continue into the measurement too:
+	 */
+	atomic_inc(&start_count);
+
+	check_tsc_warp();
+
+	while (atomic_read(&stop_count) != cpus-1)
+		cpu_relax();
+
+	/*
+	 * Reset it - just in case we boot another CPU later:
+	 */
+	atomic_set(&start_count, 0);
+
+	if (nr_warps) {
+		printk("\n");
+		printk(KERN_WARNING "Measured %Ld cycles TSC warp between CPUs,"
+				    " turning off TSC clock.\n", max_warp);
+		mark_tsc_unstable("check_tsc_sync_source failed");
+		nr_warps = 0;
+		max_warp = 0;
+		last_tsc = 0;
+	} else {
+		printk(" passed.\n");
+	}
+
+	/*
+	 * Let the target continue with the bootup:
+	 */
+	atomic_inc(&stop_count);
+}
+
+/*
+ * Freshly booted CPUs call into this:
+ */
+void __cpuinit check_tsc_sync_target(void)
+{
+	int cpus = 2;
+
+	if (unsynchronized_tsc())
+		return;
+
+	/*
+	 * Register this CPU's participation and wait for the
+	 * source CPU to start the measurement:
+	 */
+	atomic_inc(&start_count);
+	while (atomic_read(&start_count) != cpus)
+		cpu_relax();
+
+	check_tsc_warp();
+
+	/*
+	 * Ok, we are done:
+	 */
+	atomic_inc(&stop_count);
+
+	/*
+	 * Wait for the source CPU to print stuff:
+	 */
+	while (atomic_read(&stop_count) != cpus)
+		cpu_relax();
+}
+#undef NR_LOOPS
+

arch/x86/kernel/verify_cpu_64.S

@@ -0,0 +1,105 @@
+/*
+ *
+ *	verify_cpu.S - Code for cpu long mode and SSE verification. This
+ *	code has been borrowed from boot/setup.S and was introduced by
+ * 	Andi Kleen.
+ *
+ *	Copyright (c) 2007  Andi Kleen (ak@suse.de)
+ *	Copyright (c) 2007  Eric Biederman (ebiederm@xmission.com)
+ *	Copyright (c) 2007  Vivek Goyal (vgoyal@in.ibm.com)
+ *
+ * 	This source code is licensed under the GNU General Public License,
+ * 	Version 2.  See the file COPYING for more details.
+ *
+ *	This is a common code for verification whether CPU supports
+ * 	long mode and SSE or not. It is not called directly instead this
+ *	file is included at various places and compiled in that context.
+ * 	Following are the current usage.
+ *
+ * 	This file is included by both 16bit and 32bit code.
+ *
+ *	arch/x86_64/boot/setup.S : Boot cpu verification (16bit)
+ *	arch/x86_64/boot/compressed/head.S: Boot cpu verification (32bit)
+ *	arch/x86_64/kernel/trampoline.S: secondary processor verfication (16bit)
+ *	arch/x86_64/kernel/acpi/wakeup.S:Verfication at resume (16bit)
+ *
+ *	verify_cpu, returns the status of cpu check in register %eax.
+ *		0: Success    1: Failure
+ *
+ * 	The caller needs to check for the error code and take the action
+ * 	appropriately. Either display a message or halt.
+ */
+
+#include <asm/cpufeature.h>
+
+verify_cpu:
+	pushfl				# Save caller passed flags
+	pushl	$0			# Kill any dangerous flags
+	popfl
+
+	pushfl				# standard way to check for cpuid
+	popl	%eax
+	movl	%eax,%ebx
+	xorl	$0x200000,%eax
+	pushl	%eax
+	popfl
+	pushfl
+	popl	%eax
+	cmpl	%eax,%ebx
+	jz	verify_cpu_no_longmode	# cpu has no cpuid
+
+	movl	$0x0,%eax		# See if cpuid 1 is implemented
+	cpuid
+	cmpl	$0x1,%eax
+	jb	verify_cpu_no_longmode	# no cpuid 1
+
+	xor	%di,%di
+	cmpl	$0x68747541,%ebx	# AuthenticAMD
+	jnz	verify_cpu_noamd
+	cmpl	$0x69746e65,%edx
+	jnz	verify_cpu_noamd
+	cmpl	$0x444d4163,%ecx
+	jnz	verify_cpu_noamd
+	mov	$1,%di			# cpu is from AMD
+
+verify_cpu_noamd:
+	movl    $0x1,%eax		# Does the cpu have what it takes
+	cpuid
+	andl	$REQUIRED_MASK0,%edx
+	xorl	$REQUIRED_MASK0,%edx
+	jnz	verify_cpu_no_longmode
+
+	movl    $0x80000000,%eax	# See if extended cpuid is implemented
+	cpuid
+	cmpl    $0x80000001,%eax
+	jb      verify_cpu_no_longmode	# no extended cpuid
+
+	movl    $0x80000001,%eax	# Does the cpu have what it takes
+	cpuid
+	andl    $REQUIRED_MASK1,%edx
+	xorl    $REQUIRED_MASK1,%edx
+	jnz     verify_cpu_no_longmode
+
+verify_cpu_sse_test:
+	movl	$1,%eax
+	cpuid
+	andl	$SSE_MASK,%edx
+	cmpl	$SSE_MASK,%edx
+	je	verify_cpu_sse_ok
+	test	%di,%di
+	jz	verify_cpu_no_longmode	# only try to force SSE on AMD
+	movl	$0xc0010015,%ecx	# HWCR
+	rdmsr
+	btr	$15,%eax		# enable SSE
+	wrmsr
+	xor	%di,%di			# don't loop
+	jmp	verify_cpu_sse_test	# try again
+
+verify_cpu_no_longmode:
+	popfl				# Restore caller passed flags
+	movl $1,%eax
+	ret
+verify_cpu_sse_ok:
+	popfl				# Restore caller passed flags
+	xorl %eax, %eax
+	ret

arch/x86/kernel/vmlinux_64.lds.S

@@ -0,0 +1,235 @@
+/* ld script to make x86-64 Linux kernel
+ * Written by Martin Mares <mj@atrey.karlin.mff.cuni.cz>;
+ */
+
+#define LOAD_OFFSET __START_KERNEL_map
+
+#include <asm-generic/vmlinux.lds.h>
+#include <asm/page.h>
+
+#undef i386	/* in case the preprocessor is a 32bit one */
+
+OUTPUT_FORMAT("elf64-x86-64", "elf64-x86-64", "elf64-x86-64")
+OUTPUT_ARCH(i386:x86-64)
+ENTRY(phys_startup_64)
+jiffies_64 = jiffies;
+_proxy_pda = 1;
+PHDRS {
+	text PT_LOAD FLAGS(5);	/* R_E */
+	data PT_LOAD FLAGS(7);	/* RWE */
+	user PT_LOAD FLAGS(7);	/* RWE */
+	data.init PT_LOAD FLAGS(7);	/* RWE */
+	note PT_NOTE FLAGS(4);	/* R__ */
+}
+SECTIONS
+{
+  . = __START_KERNEL;
+  phys_startup_64 = startup_64 - LOAD_OFFSET;
+  _text = .;			/* Text and read-only data */
+  .text :  AT(ADDR(.text) - LOAD_OFFSET) {
+	/* First the code that has to be first for bootstrapping */
+	*(.text.head)
+	_stext = .;
+	/* Then the rest */
+	TEXT_TEXT
+	SCHED_TEXT
+	LOCK_TEXT
+	KPROBES_TEXT
+	*(.fixup)
+	*(.gnu.warning)
+	} :text = 0x9090
+  				/* out-of-line lock text */
+  .text.lock : AT(ADDR(.text.lock) - LOAD_OFFSET) { *(.text.lock) }
+
+  _etext = .;			/* End of text section */
+
+  . = ALIGN(16);		/* Exception table */
+  __start___ex_table = .;
+  __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { *(__ex_table) }
+  __stop___ex_table = .;
+
+  NOTES :text :note
+
+  BUG_TABLE :text
+
+  RODATA
+
+  . = ALIGN(4);
+  .tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) {
+  	__tracedata_start = .;
+	*(.tracedata)
+  	__tracedata_end = .;
+  }
+
+  . = ALIGN(PAGE_SIZE);        /* Align data segment to page size boundary */
+				/* Data */
+  .data : AT(ADDR(.data) - LOAD_OFFSET) {
+	DATA_DATA
+	CONSTRUCTORS
+	} :data
+
+  _edata = .;			/* End of data section */
+
+  . = ALIGN(PAGE_SIZE);
+  . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
+  .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) {
+	*(.data.cacheline_aligned)
+  }
+  . = ALIGN(CONFIG_X86_INTERNODE_CACHE_BYTES);
+  .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) {
+  	*(.data.read_mostly)
+  }
+
+#define VSYSCALL_ADDR (-10*1024*1024)
+#define VSYSCALL_PHYS_ADDR ((LOADADDR(.data.read_mostly) + SIZEOF(.data.read_mostly) + 4095) & ~(4095))
+#define VSYSCALL_VIRT_ADDR ((ADDR(.data.read_mostly) + SIZEOF(.data.read_mostly) + 4095) & ~(4095))
+
+#define VLOAD_OFFSET (VSYSCALL_ADDR - VSYSCALL_PHYS_ADDR)
+#define VLOAD(x) (ADDR(x) - VLOAD_OFFSET)
+
+#define VVIRT_OFFSET (VSYSCALL_ADDR - VSYSCALL_VIRT_ADDR)
+#define VVIRT(x) (ADDR(x) - VVIRT_OFFSET)
+
+  . = VSYSCALL_ADDR;
+  .vsyscall_0 :	 AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) } :user
+  __vsyscall_0 = VSYSCALL_VIRT_ADDR;
+
+  . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
+  .vsyscall_fn : AT(VLOAD(.vsyscall_fn)) { *(.vsyscall_fn) }
+  . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
+  .vsyscall_gtod_data : AT(VLOAD(.vsyscall_gtod_data))
+		{ *(.vsyscall_gtod_data) }
+  vsyscall_gtod_data = VVIRT(.vsyscall_gtod_data);
+  .vsyscall_clock : AT(VLOAD(.vsyscall_clock))
+		{ *(.vsyscall_clock) }
+  vsyscall_clock = VVIRT(.vsyscall_clock);
+
+
+  .vsyscall_1 ADDR(.vsyscall_0) + 1024: AT(VLOAD(.vsyscall_1))
+		{ *(.vsyscall_1) }
+  .vsyscall_2 ADDR(.vsyscall_0) + 2048: AT(VLOAD(.vsyscall_2))
+		{ *(.vsyscall_2) }
+
+  .vgetcpu_mode : AT(VLOAD(.vgetcpu_mode)) { *(.vgetcpu_mode) }
+  vgetcpu_mode = VVIRT(.vgetcpu_mode);
+
+  . = ALIGN(CONFIG_X86_L1_CACHE_BYTES);
+  .jiffies : AT(VLOAD(.jiffies)) { *(.jiffies) }
+  jiffies = VVIRT(.jiffies);
+
+  .vsyscall_3 ADDR(.vsyscall_0) + 3072: AT(VLOAD(.vsyscall_3))
+		{ *(.vsyscall_3) }
+
+  . = VSYSCALL_VIRT_ADDR + 4096;
+
+#undef VSYSCALL_ADDR
+#undef VSYSCALL_PHYS_ADDR
+#undef VSYSCALL_VIRT_ADDR
+#undef VLOAD_OFFSET
+#undef VLOAD
+#undef VVIRT_OFFSET
+#undef VVIRT
+
+  . = ALIGN(8192);		/* init_task */
+  .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) {
+	*(.data.init_task)
+  }:data.init
+
+  . = ALIGN(4096);
+  .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) {
+	*(.data.page_aligned)
+  }
+
+  /* might get freed after init */
+  . = ALIGN(4096);
+  __smp_alt_begin = .;
+  __smp_locks = .;
+  .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) {
+	*(.smp_locks)
+  }
+  __smp_locks_end = .;
+  . = ALIGN(4096);
+  __smp_alt_end = .;
+
+  . = ALIGN(4096);		/* Init code and data */
+  __init_begin = .;
+  .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) {
+	_sinittext = .;
+	*(.init.text)
+	_einittext = .;
+  }
+  __initdata_begin = .;
+  .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { *(.init.data) }
+  __initdata_end = .;
+  . = ALIGN(16);
+  __setup_start = .;
+  .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { *(.init.setup) }
+  __setup_end = .;
+  __initcall_start = .;
+  .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) {
+	INITCALLS
+  }
+  __initcall_end = .;
+  __con_initcall_start = .;
+  .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) {
+	*(.con_initcall.init)
+  }
+  __con_initcall_end = .;
+  SECURITY_INIT
+  . = ALIGN(8);
+  __alt_instructions = .;
+  .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) {
+	*(.altinstructions)
+  }
+  __alt_instructions_end = .; 
+  .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) {
+	*(.altinstr_replacement)
+  }
+  /* .exit.text is discard at runtime, not link time, to deal with references
+     from .altinstructions and .eh_frame */
+  .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { *(.exit.text) }
+  .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { *(.exit.data) }
+
+/* vdso blob that is mapped into user space */
+  vdso_start = . ;
+  .vdso  : AT(ADDR(.vdso) - LOAD_OFFSET) { *(.vdso) }
+  . = ALIGN(4096);
+  vdso_end = .;
+
+#ifdef CONFIG_BLK_DEV_INITRD
+  . = ALIGN(4096);
+  __initramfs_start = .;
+  .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { *(.init.ramfs) }
+  __initramfs_end = .;
+#endif
+
+  PERCPU(4096)
+
+  . = ALIGN(4096);
+  __init_end = .;
+
+  . = ALIGN(4096);
+  __nosave_begin = .;
+  .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { *(.data.nosave) }
+  . = ALIGN(4096);
+  __nosave_end = .;
+
+  __bss_start = .;		/* BSS */
+  .bss : AT(ADDR(.bss) - LOAD_OFFSET) {
+	*(.bss.page_aligned)
+	*(.bss)
+	}
+  __bss_stop = .;
+
+  _end = . ;
+
+  /* Sections to be discarded */
+  /DISCARD/ : {
+	*(.exitcall.exit)
+	*(.eh_frame)
+	}
+
+  STABS_DEBUG
+
+  DWARF_DEBUG
+}

arch/x86/kernel/vsmp_64.c

@@ -0,0 +1,49 @@
+/*
+ * vSMPowered(tm) systems specific initialization
+ * Copyright (C) 2005 ScaleMP Inc.
+ *
+ * Use of this code is subject to the terms and conditions of the
+ * GNU general public license version 2. See "COPYING" or
+ * http://www.gnu.org/licenses/gpl.html
+ *
+ * Ravikiran Thirumalai <kiran@scalemp.com>,
+ * Shai Fultheim <shai@scalemp.com>
+ */
+
+#include <linux/init.h>
+#include <linux/pci_ids.h>
+#include <linux/pci_regs.h>
+#include <asm/pci-direct.h>
+#include <asm/io.h>
+
+static int __init vsmp_init(void)
+{
+	void *address;
+	unsigned int cap, ctl;
+
+	if (!early_pci_allowed())
+		return 0;
+
+	/* Check if we are running on a ScaleMP vSMP box */
+	if ((read_pci_config_16(0, 0x1f, 0, PCI_VENDOR_ID) != PCI_VENDOR_ID_SCALEMP) ||
+	    (read_pci_config_16(0, 0x1f, 0, PCI_DEVICE_ID) != PCI_DEVICE_ID_SCALEMP_VSMP_CTL))
+		return 0;
+
+	/* set vSMP magic bits to indicate vSMP capable kernel */
+	address = ioremap(read_pci_config(0, 0x1f, 0, PCI_BASE_ADDRESS_0), 8);
+	cap = readl(address);
+	ctl = readl(address + 4);
+	printk("vSMP CTL: capabilities:0x%08x  control:0x%08x\n", cap, ctl);
+	if (cap & ctl & (1 << 4)) {
+		/* Turn on vSMP IRQ fastpath handling (see system.h) */
+		ctl &= ~(1 << 4);
+		writel(ctl, address + 4);
+		ctl = readl(address + 4);
+		printk("vSMP CTL: control set to:0x%08x\n", ctl);
+	}
+
+	iounmap(address);
+	return 0;
+}
+
+core_initcall(vsmp_init);

arch/x86/kernel/vsyscall_64.c

@@ -0,0 +1,349 @@
+/*
+ *  linux/arch/x86_64/kernel/vsyscall.c
+ *
+ *  Copyright (C) 2001 Andrea Arcangeli <andrea@suse.de> SuSE
+ *  Copyright 2003 Andi Kleen, SuSE Labs.
+ *
+ *  Thanks to hpa@transmeta.com for some useful hint.
+ *  Special thanks to Ingo Molnar for his early experience with
+ *  a different vsyscall implementation for Linux/IA32 and for the name.
+ *
+ *  vsyscall 1 is located at -10Mbyte, vsyscall 2 is located
+ *  at virtual address -10Mbyte+1024bytes etc... There are at max 4
+ *  vsyscalls. One vsyscall can reserve more than 1 slot to avoid
+ *  jumping out of line if necessary. We cannot add more with this
+ *  mechanism because older kernels won't return -ENOSYS.
+ *  If we want more than four we need a vDSO.
+ *
+ *  Note: the concept clashes with user mode linux. If you use UML and
+ *  want per guest time just set the kernel.vsyscall64 sysctl to 0.
+ */
+
+#include <linux/time.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/timer.h>
+#include <linux/seqlock.h>
+#include <linux/jiffies.h>
+#include <linux/sysctl.h>
+#include <linux/clocksource.h>
+#include <linux/getcpu.h>
+#include <linux/cpu.h>
+#include <linux/smp.h>
+#include <linux/notifier.h>
+
+#include <asm/vsyscall.h>
+#include <asm/pgtable.h>
+#include <asm/page.h>
+#include <asm/unistd.h>
+#include <asm/fixmap.h>
+#include <asm/errno.h>
+#include <asm/io.h>
+#include <asm/segment.h>
+#include <asm/desc.h>
+#include <asm/topology.h>
+#include <asm/vgtod.h>
+
+#define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr)))
+#define __syscall_clobber "r11","rcx","memory"
+#define __pa_vsymbol(x)			\
+	({unsigned long v;  		\
+	extern char __vsyscall_0; 	\
+	  asm("" : "=r" (v) : "0" (x)); \
+	  ((v - VSYSCALL_FIRST_PAGE) + __pa_symbol(&__vsyscall_0)); })
+
+/*
+ * vsyscall_gtod_data contains data that is :
+ * - readonly from vsyscalls
+ * - writen by timer interrupt or systcl (/proc/sys/kernel/vsyscall64)
+ * Try to keep this structure as small as possible to avoid cache line ping pongs
+ */
+int __vgetcpu_mode __section_vgetcpu_mode;
+
+struct vsyscall_gtod_data __vsyscall_gtod_data __section_vsyscall_gtod_data =
+{
+	.lock = SEQLOCK_UNLOCKED,
+	.sysctl_enabled = 1,
+};
+
+void update_vsyscall(struct timespec *wall_time, struct clocksource *clock)
+{
+	unsigned long flags;
+
+	write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
+	/* copy vsyscall data */
+	vsyscall_gtod_data.clock.vread = clock->vread;
+	vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
+	vsyscall_gtod_data.clock.mask = clock->mask;
+	vsyscall_gtod_data.clock.mult = clock->mult;
+	vsyscall_gtod_data.clock.shift = clock->shift;
+	vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec;
+	vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
+	vsyscall_gtod_data.sys_tz = sys_tz;
+	vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
+	vsyscall_gtod_data.wall_to_monotonic = wall_to_monotonic;
+	write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
+}
+
+/* RED-PEN may want to readd seq locking, but then the variable should be
+ * write-once.
+ */
+static __always_inline void do_get_tz(struct timezone * tz)
+{
+	*tz = __vsyscall_gtod_data.sys_tz;
+}
+
+static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz)
+{
+	int ret;
+	asm volatile("vsysc2: syscall"
+		: "=a" (ret)
+		: "0" (__NR_gettimeofday),"D" (tv),"S" (tz)
+		: __syscall_clobber );
+	return ret;
+}
+
+static __always_inline long time_syscall(long *t)
+{
+	long secs;
+	asm volatile("vsysc1: syscall"
+		: "=a" (secs)
+		: "0" (__NR_time),"D" (t) : __syscall_clobber);
+	return secs;
+}
+
+static __always_inline void do_vgettimeofday(struct timeval * tv)
+{
+	cycle_t now, base, mask, cycle_delta;
+	unsigned seq;
+	unsigned long mult, shift, nsec;
+	cycle_t (*vread)(void);
+	do {
+		seq = read_seqbegin(&__vsyscall_gtod_data.lock);
+
+		vread = __vsyscall_gtod_data.clock.vread;
+		if (unlikely(!__vsyscall_gtod_data.sysctl_enabled || !vread)) {
+			gettimeofday(tv,NULL);
+			return;
+		}
+		now = vread();
+		base = __vsyscall_gtod_data.clock.cycle_last;
+		mask = __vsyscall_gtod_data.clock.mask;
+		mult = __vsyscall_gtod_data.clock.mult;
+		shift = __vsyscall_gtod_data.clock.shift;
+
+		tv->tv_sec = __vsyscall_gtod_data.wall_time_sec;
+		nsec = __vsyscall_gtod_data.wall_time_nsec;
+	} while (read_seqretry(&__vsyscall_gtod_data.lock, seq));
+
+	/* calculate interval: */
+	cycle_delta = (now - base) & mask;
+	/* convert to nsecs: */
+	nsec += (cycle_delta * mult) >> shift;
+
+	while (nsec >= NSEC_PER_SEC) {
+		tv->tv_sec += 1;
+		nsec -= NSEC_PER_SEC;
+	}
+	tv->tv_usec = nsec / NSEC_PER_USEC;
+}
+
+int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
+{
+	if (tv)
+		do_vgettimeofday(tv);
+	if (tz)
+		do_get_tz(tz);
+	return 0;
+}
+
+/* This will break when the xtime seconds get inaccurate, but that is
+ * unlikely */
+time_t __vsyscall(1) vtime(time_t *t)
+{
+	struct timeval tv;
+	time_t result;
+	if (unlikely(!__vsyscall_gtod_data.sysctl_enabled))
+		return time_syscall(t);
+
+	vgettimeofday(&tv, 0);
+	result = tv.tv_sec;
+	if (t)
+		*t = result;
+	return result;
+}
+
+/* Fast way to get current CPU and node.
+   This helps to do per node and per CPU caches in user space.
+   The result is not guaranteed without CPU affinity, but usually
+   works out because the scheduler tries to keep a thread on the same
+   CPU.
+
+   tcache must point to a two element sized long array.
+   All arguments can be NULL. */
+long __vsyscall(2)
+vgetcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache)
+{
+	unsigned int dummy, p;
+	unsigned long j = 0;
+
+	/* Fast cache - only recompute value once per jiffies and avoid
+	   relatively costly rdtscp/cpuid otherwise.
+	   This works because the scheduler usually keeps the process
+	   on the same CPU and this syscall doesn't guarantee its
+	   results anyways.
+	   We do this here because otherwise user space would do it on
+	   its own in a likely inferior way (no access to jiffies).
+	   If you don't like it pass NULL. */
+	if (tcache && tcache->blob[0] == (j = __jiffies)) {
+		p = tcache->blob[1];
+	} else if (__vgetcpu_mode == VGETCPU_RDTSCP) {
+		/* Load per CPU data from RDTSCP */
+		rdtscp(dummy, dummy, p);
+	} else {
+		/* Load per CPU data from GDT */
+		asm("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG));
+	}
+	if (tcache) {
+		tcache->blob[0] = j;
+		tcache->blob[1] = p;
+	}
+	if (cpu)
+		*cpu = p & 0xfff;
+	if (node)
+		*node = p >> 12;
+	return 0;
+}
+
+long __vsyscall(3) venosys_1(void)
+{
+	return -ENOSYS;
+}
+
+#ifdef CONFIG_SYSCTL
+
+#define SYSCALL 0x050f
+#define NOP2    0x9090
+
+/*
+ * NOP out syscall in vsyscall page when not needed.
+ */
+static int vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp,
+                        void __user *buffer, size_t *lenp, loff_t *ppos)
+{
+	extern u16 vsysc1, vsysc2;
+	u16 __iomem *map1;
+	u16 __iomem *map2;
+	int ret = proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
+	if (!write)
+		return ret;
+	/* gcc has some trouble with __va(__pa()), so just do it this
+	   way. */
+	map1 = ioremap(__pa_vsymbol(&vsysc1), 2);
+	if (!map1)
+		return -ENOMEM;
+	map2 = ioremap(__pa_vsymbol(&vsysc2), 2);
+	if (!map2) {
+		ret = -ENOMEM;
+		goto out;
+	}
+	if (!vsyscall_gtod_data.sysctl_enabled) {
+		writew(SYSCALL, map1);
+		writew(SYSCALL, map2);
+	} else {
+		writew(NOP2, map1);
+		writew(NOP2, map2);
+	}
+	iounmap(map2);
+out:
+	iounmap(map1);
+	return ret;
+}
+
+static int vsyscall_sysctl_nostrat(ctl_table *t, int __user *name, int nlen,
+				void __user *oldval, size_t __user *oldlenp,
+				void __user *newval, size_t newlen)
+{
+	return -ENOSYS;
+}
+
+static ctl_table kernel_table2[] = {
+	{ .ctl_name = 99, .procname = "vsyscall64",
+	  .data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int),
+	  .mode = 0644,
+	  .strategy = vsyscall_sysctl_nostrat,
+	  .proc_handler = vsyscall_sysctl_change },
+	{}
+};
+
+static ctl_table kernel_root_table2[] = {
+	{ .ctl_name = CTL_KERN, .procname = "kernel", .mode = 0555,
+	  .child = kernel_table2 },
+	{}
+};
+
+#endif
+
+/* Assume __initcall executes before all user space. Hopefully kmod
+   doesn't violate that. We'll find out if it does. */
+static void __cpuinit vsyscall_set_cpu(int cpu)
+{
+	unsigned long *d;
+	unsigned long node = 0;
+#ifdef CONFIG_NUMA
+	node = cpu_to_node[cpu];
+#endif
+	if (cpu_has(&cpu_data[cpu], X86_FEATURE_RDTSCP))
+		write_rdtscp_aux((node << 12) | cpu);
+
+	/* Store cpu number in limit so that it can be loaded quickly
+	   in user space in vgetcpu.
+	   12 bits for the CPU and 8 bits for the node. */
+	d = (unsigned long *)(cpu_gdt(cpu) + GDT_ENTRY_PER_CPU);
+	*d = 0x0f40000000000ULL;
+	*d |= cpu;
+	*d |= (node & 0xf) << 12;
+	*d |= (node >> 4) << 48;
+}
+
+static void __cpuinit cpu_vsyscall_init(void *arg)
+{
+	/* preemption should be already off */
+	vsyscall_set_cpu(raw_smp_processor_id());
+}
+
+static int __cpuinit
+cpu_vsyscall_notifier(struct notifier_block *n, unsigned long action, void *arg)
+{
+	long cpu = (long)arg;
+	if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN)
+		smp_call_function_single(cpu, cpu_vsyscall_init, NULL, 0, 1);
+	return NOTIFY_DONE;
+}
+
+static void __init map_vsyscall(void)
+{
+	extern char __vsyscall_0;
+	unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);
+
+	/* Note that VSYSCALL_MAPPED_PAGES must agree with the code below. */
+	__set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
+}
+
+static int __init vsyscall_init(void)
+{
+	BUG_ON(((unsigned long) &vgettimeofday !=
+			VSYSCALL_ADDR(__NR_vgettimeofday)));
+	BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime));
+	BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
+	BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu));
+	map_vsyscall();
+#ifdef CONFIG_SYSCTL
+	register_sysctl_table(kernel_root_table2);
+#endif
+	on_each_cpu(cpu_vsyscall_init, NULL, 0, 1);
+	hotcpu_notifier(cpu_vsyscall_notifier, 0);
+	return 0;
+}
+
+__initcall(vsyscall_init);

arch/x86/kernel/x8664_ksyms_64.c

@@ -0,0 +1,62 @@
+/* Exports for assembly files.
+   All C exports should go in the respective C files. */
+
+#include <linux/module.h>
+#include <linux/smp.h>
+
+#include <asm/semaphore.h>
+#include <asm/processor.h>
+#include <asm/uaccess.h>
+#include <asm/pgtable.h>
+
+EXPORT_SYMBOL(kernel_thread);
+
+EXPORT_SYMBOL(__down_failed);
+EXPORT_SYMBOL(__down_failed_interruptible);
+EXPORT_SYMBOL(__down_failed_trylock);
+EXPORT_SYMBOL(__up_wakeup);
+
+EXPORT_SYMBOL(__get_user_1);
+EXPORT_SYMBOL(__get_user_2);
+EXPORT_SYMBOL(__get_user_4);
+EXPORT_SYMBOL(__get_user_8);
+EXPORT_SYMBOL(__put_user_1);
+EXPORT_SYMBOL(__put_user_2);
+EXPORT_SYMBOL(__put_user_4);
+EXPORT_SYMBOL(__put_user_8);
+
+EXPORT_SYMBOL(copy_user_generic);
+EXPORT_SYMBOL(__copy_user_nocache);
+EXPORT_SYMBOL(copy_from_user);
+EXPORT_SYMBOL(copy_to_user);
+EXPORT_SYMBOL(__copy_from_user_inatomic);
+
+EXPORT_SYMBOL(copy_page);
+EXPORT_SYMBOL(clear_page);
+
+#ifdef CONFIG_SMP
+extern void  __write_lock_failed(rwlock_t *rw);
+extern void  __read_lock_failed(rwlock_t *rw);
+EXPORT_SYMBOL(__write_lock_failed);
+EXPORT_SYMBOL(__read_lock_failed);
+#endif
+
+/* Export string functions. We normally rely on gcc builtin for most of these,
+   but gcc sometimes decides not to inline them. */    
+#undef memcpy
+#undef memset
+#undef memmove
+
+extern void * memset(void *,int,__kernel_size_t);
+extern void * memcpy(void *,const void *,__kernel_size_t);
+extern void * __memcpy(void *,const void *,__kernel_size_t);
+
+EXPORT_SYMBOL(memset);
+EXPORT_SYMBOL(memcpy);
+EXPORT_SYMBOL(__memcpy);
+
+EXPORT_SYMBOL(empty_zero_page);
+EXPORT_SYMBOL(init_level4_pgt);
+EXPORT_SYMBOL(load_gs_index);
+
+EXPORT_SYMBOL(_proxy_pda);