[PATCH] x86-64 merge

Only bug fixes and making it compile again and a few minor features.
Also one security fix that got lost earlier

- Document boot options
- Better cpu local data
- Emulate FIOQSIZE
- Fix return value of 32bit ipccall
- Various minor style fixes
- Save some memory in apic tables
- Merge with 2.6.0test2/i386
- Readd ioport fix
- Sort exception tables at boot time
- Add local.h
- Fix for_each_cpu on UP
- Add utimes and tgkill system calls for 64bit
- Update defconfig

Documentation/x86_64/boot-options.txt

+AMD64 specific boot options
+
+There are many others (usually documented in driver documentation), but
+only the AMD64 specific ones are listed here.
+
+Machine check
+
+(see the Opteron BIOS&Kernel manual for more details on the banks etc.)
+
+   mce=off disable machine check
+   mce=nok8 disable k8 specific features
+   mce=disable<NUMBER> disable bank NUMBER
+   mce=enable<NUMBER> enable bank number
+   mce=device	Enable more machine check options in Northbridge.
+		Can be useful for device driver debugging.
+   mce=NUMBER mcheck timer interval number seconds.
+   Can be also comma separated in a single mce=
+
+   nomce (for compatibility with i386): same as mce=off
+
+APICs
+
+   nolocalapic   Don't use a local or IO-APIC. This should only
+   		 be needed if you have a buggy BIOS. The newer
+		 kernels already turn it off by default if the
+		 BIOS didn't enable the local APIC, so it will
+		 be hopefully not needed.
+		 Note this code path is not very well tested, you are on
+		 your own.
+
+   apic		 Use IO-APIC. Default
+
+   noapic	 Don't use the IO-APIC.
+		 Also only lightly tested.
+
+   pirq=...	 See Documentation/i386/IO-APIC.txt
+
+Early Console
+
+   syntax: earlyprintk=vga
+           earlyprintk=serial[,ttySn[,baudrate]]
+
+   The early console is useful when the kernel crashes before the
+   normal console is initialized. It is not enabled by
+   default because it has some cosmetic problems.
+   Append ,keep to not disable it when the real console takes over.
+   Only vga or serial at a time, not both.
+   Currently only ttyS0 and ttyS1 are supported.
+   Interaction with the standard serial driver is not very good.
+   The VGA output is eventually overwritten by the real console.
+
+Timing
+
+  notsc
+  Don't use the CPU time stamp counter to read the wall time.
+  This can be used to work around timing problems on multiprocessor systems
+  with not properly synchronized CPUs. Only useful with a SMP kernel
+
+  report_lost_ticks
+  Report when timer interrupts are lost because some code turned off
+  interrupts for too long.
+
+  nmi_watchdog=NUMBER
+  NUMBER can be:
+  0 don't use an NMI watchdog
+  1 use the IO-APIC timer for the NMI watchdog
+  2 use the local APIC for the NMI watchdog using a performance counter. Note
+  This will use one performance counter and the local APIC's performance
+  vector.
+
+Idle loop
+
+  idle=poll
+  Don't do power saving in the idle loop using HLT, but poll for rescheduling
+  event. This will make the CPUs eat a lot more power, but may be useful
+  to get slightly better performance in multiprocessor benchmarks. It also
+  makes some profiling using performance counters more accurate.
+
+Rebooting
+
+   reboot=b[ios] | t[riple] | k[bd] [, [w]arm | [c]old]
+   bios	  Use the CPU reboto vector for warm reset
+   warm   Don't set the cold reboot flag
+   cold   Set the cold reboto flag
+   triple Force a triple fault (init)
+   kbd    Use the keyboard controller. cold reset (default)
+
+   Using warm reset will be much faster especially on big memory
+   systems because the BIOS will not go through the memory check.
+   Disadvantage is that not all hardware will be completely reinitialized
+   on reboot so there may be boot problems on some systems.
+
+Non Executable Mappings
+
+  noexec=on|off
+
+  on      Enable
+  off     Disable
+  noforce (default) Don't enable by default for heap/stack/data,
+          but allow PROT_EXEC to be effective
+
+  noexec32=opt{,opt}
+
+  Control the no exec default for 32bit processes.
+  Requires noexec=on or noexec=noforce to be effective.
+
+  Valid options:
+     all,on    Heap,stack,data is non executable.
+     off       (default) Heap,stack,data is executable
+     stack     Stack is non executable, heap/data is.
+     force     Don't imply PROT_EXEC for PROT_READ
+     compat    (default) Imply PROT_EXEC for PROT_READ
+
+SMP
+
+  nosmp	Only use a single CPU
+
+  maxcpus=NUMBER only use upto NUMBER CPUs
+
+  cpumask=MASK   only use cpus with bits set in mask
+
+NUMA
+
+  numa=off	Only set up a single NUMA node spanning all memory.
+
+
+ACPI
+
+  acpi=off	Don't enable ACPI
+
+PCI
+
+  pci=off	Don't use PCI
+  pci=conf1	Use conf1 access.
+  pci=conf2	Use conf2 access.
+  pci=rom	Assign ROMs.
+  pci=assign-busses    Assign busses
+  pci=irqmask=MASK	       Set PCI interrupt mask to MASK
+  pci=lastbus=NUMBER	       Scan upto NUMBER busses, no matter what the mptable says.
+
+IOMMU
+
+  iommu=[size][,noagp][,off][,force][,noforce][,leak][,memaper[=order]]
+   size  set size of iommu (in bytes)
+   noagp don't initialize the AGP driver and use full aperture.
+   off   don't use the IOMMU
+   leak  turn on simple iommu leak tracing (only when CONFIG_IOMMU_LEAK is on)
+   memaper[=order] allocate an own aperture over RAM with size 32MB^order.
+   noforce don't force IOMMU usage. Default.
+   force  Force IOMMU
+
+

arch/x86_64/defconfig

@@ -24,8 +24,11 @@ CONFIG_SYSVIPC=y
 CONFIG_SYSCTL=y
 CONFIG_LOG_BUF_SHIFT=16
 # CONFIG_EMBEDDED is not set
+CONFIG_KALLSYMS=y
 CONFIG_FUTEX=y
 CONFIG_EPOLL=y
+CONFIG_IOSCHED_AS=y
+CONFIG_IOSCHED_DEADLINE=y
 
 #
 # Loadable module support
@@ -135,6 +138,7 @@ CONFIG_BLK_DEV_FD=y
 # CONFIG_BLK_DEV_DAC960 is not set
 # CONFIG_BLK_DEV_UMEM is not set
 CONFIG_BLK_DEV_LOOP=y
+# CONFIG_BLK_DEV_CRYPTOLOOP is not set
 # CONFIG_BLK_DEV_NBD is not set
 CONFIG_BLK_DEV_RAM=y
 CONFIG_BLK_DEV_RAM_SIZE=4096
@@ -163,6 +167,7 @@ CONFIG_BLK_DEV_IDECD=y
 # CONFIG_BLK_DEV_IDEFLOPPY is not set
 # CONFIG_BLK_DEV_IDESCSI is not set
 # CONFIG_IDE_TASK_IOCTL is not set
+# CONFIG_IDE_TASKFILE_IO is not set
 
 #
 # IDE chipset support/bugfixes
@@ -322,7 +327,7 @@ CONFIG_IPV6_PRIVACY=y
 # CONFIG_XFRM_USER is not set
 
 #
-#    SCTP Configuration (EXPERIMENTAL)
+# SCTP Configuration (EXPERIMENTAL)
 #
 CONFIG_IPV6_SCTP__=y
 # CONFIG_IP_SCTP is not set
@@ -379,7 +384,6 @@ CONFIG_NET_PCI=y
 CONFIG_AMD8111_ETH=y
 # CONFIG_ADAPTEC_STARFIRE is not set
 # CONFIG_B44 is not set
-# CONFIG_TC35815 is not set
 # CONFIG_DGRS is not set
 # CONFIG_EEPRO100 is not set
 # CONFIG_E100 is not set
@@ -727,6 +731,9 @@ CONFIG_SOUND_ICH=y
 # CONFIG_SOUND_MSNDPIN is not set
 # CONFIG_SOUND_VIA82CXXX is not set
 # CONFIG_SOUND_OSS is not set
+# CONFIG_SOUND_ALI5455 is not set
+# CONFIG_SOUND_FORTE is not set
+# CONFIG_SOUND_AD1980 is not set
 
 #
 # USB support
@@ -753,10 +760,8 @@ CONFIG_DEBUG_KERNEL=y
 CONFIG_MAGIC_SYSRQ=y
 # CONFIG_DEBUG_SPINLOCK is not set
 # CONFIG_INIT_DEBUG is not set
-CONFIG_KALLSYMS=y
 # CONFIG_FRAME_POINTER is not set
-CONFIG_IOMMU_DEBUG=y
-CONFIG_IOMMU_LEAK=y
+# CONFIG_IOMMU_DEBUG is not set
 CONFIG_MCE_DEBUG=y
 
 #

arch/x86_64/ia32/ia32_binfmt.c

@@ -32,12 +32,14 @@
 #define AT_SYSINFO 32
 #define AT_SYSINFO_EHDR		33
 
-#if 0 /* disabled for now because the code has still problems */
+int sysctl_vsyscall32;
+
 #define ARCH_DLINFO do {  \
+	if (sysctl_vsyscall32) { \
 	NEW_AUX_ENT(AT_SYSINFO, (u32)(u64)VSYSCALL32_VSYSCALL); \
 	NEW_AUX_ENT(AT_SYSINFO_EHDR, VSYSCALL32_BASE);    \
+	}	\
 } while(0)
-#endif
 
 struct file;
 struct elf_phdr; 
@@ -202,7 +204,7 @@ static inline int elf_core_copy_task_regs(struct task_struct *t, elf_gregset_t*
 }
 
 static inline int 
-elf_core_copy_task_fpregs(struct task_struct *tsk, elf_fpregset_t *fpu)
+elf_core_copy_task_fpregs(struct task_struct *tsk, struct pt_regs *xregs, elf_fpregset_t *fpu)
 {
 	struct _fpstate_ia32 *fpstate = (void*)fpu; 
 	struct pt_regs *regs = (struct pt_regs *)(tsk->thread.rsp0); 

arch/x86_64/ia32/ia32_ioctl.c

@@ -726,6 +726,7 @@ COMPATIBLE_IOCTL(BNEPCONNADD)
 COMPATIBLE_IOCTL(BNEPCONNDEL)
 COMPATIBLE_IOCTL(BNEPGETCONNLIST)
 COMPATIBLE_IOCTL(BNEPGETCONNINFO)
+COMPATIBLE_IOCTL(FIOQSIZE)
 
 /* And these ioctls need translation */
 HANDLE_IOCTL(TIOCGDEV, tiocgdev)

arch/x86_64/ia32/ipc32.c

@@ -687,9 +687,7 @@ sys32_ipc (u32 call, int first, int second, int third, u32 ptr, u32 fifth)
 		return sys_shmget(first, second, third);
 	      case SHMCTL:
 		return shmctl32(first, second, (void *)AA(ptr));
-	      default:
-		return -EINVAL;
 	}
-	return -EINVAL;
+	return -ENOSYS;
 }
 

arch/x86_64/kernel/apic.c

@@ -441,9 +441,6 @@ void __init setup_local_APIC (void)
 
 #ifdef CONFIG_PM
 
-#include <linux/device.h>
-#include <linux/module.h>
-
 static struct {
 	/* 'active' is true if the local APIC was enabled by us and
 	   not the BIOS; this signifies that we are also responsible
@@ -540,7 +537,6 @@ static struct sysdev_class lapic_sysclass = {
 	.suspend	= lapic_suspend,
 };
 
-/* not static, needed by child devices */
 static struct sys_device device_lapic = {
 	.id		= 0,
 	.cls		= &lapic_sysclass,

arch/x86_64/kernel/bluesmoke.c

@@ -127,9 +127,8 @@ static struct pci_dev *find_k8_nb(void)
 { 
 	struct pci_dev *dev = NULL;
 	int cpu = smp_processor_id(); 
-	while ((dev = pci_find_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
-		if (dev->bus->number==0 && PCI_FUNC(dev->devfn)==3 &&
-		    PCI_SLOT(dev->devfn) == (24U+cpu))
+	while ((dev = pci_find_device(PCI_VENDOR_ID_AMD, 0x1103, dev)) != NULL) {
+		if (dev->bus->number==0 && PCI_SLOT(dev->devfn) == (24U+cpu))
 			return dev;
 	}
 	return NULL;
@@ -198,7 +197,7 @@ static char *highbits[32] = {
 	[0] = "err cpu1",
 };
 
-static void check_k8_nb(void)
+static void check_k8_nb(int header)
 {
 	struct pci_dev *nb;
 	nb = find_k8_nb(); 
@@ -210,6 +209,9 @@ static void check_k8_nb(void)
 	pci_read_config_dword(nb, 0x4c, &statushigh);
 	if (!(statushigh & (1<<31)))
 		return;
+	if (header)
+		printk(KERN_ERR "CPU %d: Silent Northbridge MCE\n", smp_processor_id());
+
 	printk(KERN_ERR "Northbridge status %08x%08x\n",
 	       statushigh,statuslow); 
 
@@ -271,9 +273,12 @@ static void k8_machine_check(struct pt_regs * regs, long error_code)
 	rdmsrl(MSR_IA32_MCG_STATUS, status); 
 	if ((status & (1<<2)) == 0) { 
 		if (!regs) 
-			check_k8_nb();
+			check_k8_nb(1);
 		return; 
 		}
+
+	printk(KERN_EMERG "CPU %d: Machine Check Exception: %016Lx\n", smp_processor_id(), status);
+
 	if (status & 1)
 		printk(KERN_EMERG "MCG_STATUS: unrecoverable\n"); 
 
@@ -291,7 +296,7 @@ static void k8_machine_check(struct pt_regs * regs, long error_code)
 	if (nbstatus & (1UL<57))
 		printk(KERN_EMERG "Unrecoverable condition\n"); 
 		
-	check_k8_nb();
+	check_k8_nb(0);
 
 	if (nbstatus & (1UL<<58)) { 
 		u64 adr;

arch/x86_64/kernel/head.S

@@ -307,7 +307,8 @@ ENTRY(empty_bad_pmd_table)
 ENTRY(level3_physmem_pgt)
 	.quad	0x0000000000105007		/* -> level2_kernel_pgt (so that __va works even before pagetable_init) */
 
-.org 0xb000
+	.org 0xb000
+#ifdef CONFIG_ACPI_SLEEP
 ENTRY(wakeup_level4_pgt)
 	.quad	0x0000000000102007		/* -> level3_ident_pgt */
 	.fill	255,8,0
@@ -315,9 +316,9 @@ ENTRY(wakeup_level4_pgt)
 	.fill	254,8,0
 	/* (2^48-(2*1024*1024*1024))/(2^39) = 511 */
 	.quad	0x0000000000103007		/* -> level3_kernel_pgt */
+#endif
 
-.org 0xc000
-.data
+	.data
 
 	.align 16
 	.globl cpu_gdt_descr

arch/x86_64/kernel/io_apic.c

@@ -63,7 +63,7 @@ int nr_ioapic_registers[MAX_IO_APICS];
  */
 
 static struct irq_pin_list {
-	int apic, pin, next;
+	short apic, pin, next;
 } irq_2_pin[PIN_MAP_SIZE];
 
 /*
@@ -1781,3 +1781,21 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq)
 }
 
 #endif /*CONFIG_ACPI_BOOT*/
+
+#ifndef CONFIG_SMP
+void send_IPI_self(int vector)
+{
+	unsigned int cfg;
+
+       /*
+        * Wait for idle.
+        */
+	apic_wait_icr_idle();
+	cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL;
+
+	/*
+	 * Send the IPI. The write to APIC_ICR fires this off.
+	 */
+	apic_write_around(APIC_ICR, cfg);
+}
+#endif

arch/x86_64/kernel/ioport.c

@@ -51,42 +51,50 @@ static void set_bitmap(unsigned long *bitmap, short base, short extent, int new_
 	}
 }
 
+
 /*
  * this changes the io permissions bitmap in the current task.
  */
 asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on)
 {
 	struct thread_struct * t = &current->thread;
-	struct tss_struct * tss;
-	int ret = 0;
+	int cpu = get_cpu();
 
 	if ((from + num <= from) || (from + num > IO_BITMAP_SIZE*32))
 		return -EINVAL;
 	if (turn_on && !capable(CAP_SYS_RAWIO))
 		return -EPERM;
 
+	struct tss_struct * tss = init_tss + cpu;
+
+	/*
+	 * 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) { 
-		t->io_bitmap_ptr = kmalloc((IO_BITMAP_SIZE+1)*4, GFP_KERNEL);
+		t->io_bitmap_ptr = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
 		if (!t->io_bitmap_ptr) { 
-			ret = -ENOMEM;
-			goto out;
-		}
-		memset(t->io_bitmap_ptr,0xff,(IO_BITMAP_SIZE+1)*4);
-		tss  = init_tss + get_cpu();
-		tss->io_map_base = IO_BITMAP_OFFSET;
 		put_cpu(); 
+			return -ENOMEM;
+		}
+
+		memset(t->io_bitmap_ptr,0xff,IO_BITMAP_BYTES);
 	}
-	tss = init_tss + get_cpu();
 
 	/*
 	 * do it in the per-thread copy and in the TSS ...
 	 */
 	set_bitmap((unsigned long *) t->io_bitmap_ptr, from, num, !turn_on);
+	if (tss->io_map_base != IO_BITMAP_OFFSET) {
+		memcpy(tss->io_bitmap, t->io_bitmap_ptr, sizeof(tss->io_bitmap));
+		tss->io_map_base = IO_BITMAP_OFFSET;
+	} else {
 	set_bitmap((unsigned long *) tss->io_bitmap, from, num, !turn_on);
+	}
 
- out:
 	put_cpu();
-	return ret;
+	return 0;
 }
 
 /*

arch/x86_64/kernel/mpparse.c

@@ -41,8 +41,10 @@ int acpi_found_madt;
  * MP-table.
  */
 int apic_version [MAX_APICS];
-int mp_bus_id_to_type [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
+unsigned char mp_bus_id_to_type [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
 int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 };
+unsigned long mp_bus_to_cpumask [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1UL };
+
 int mp_current_pci_id = 0;
 /* I/O APIC entries */
 struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];

arch/x86_64/kernel/nmi.c

@@ -141,14 +141,14 @@ void disable_lapic_nmi_watchdog(void)
 	/* tell do_nmi() and others that we're not active any more */
 	nmi_watchdog = 0;
 }
+
 void enable_lapic_nmi_watchdog(void)
-  {
+{
 	if (nmi_active < 0) {
 		nmi_watchdog = NMI_LOCAL_APIC;
 		setup_apic_nmi_watchdog();
 	}
-  }
-
+}
 
 void disable_timer_nmi_watchdog(void)
 {
@@ -173,8 +173,6 @@ void enable_timer_nmi_watchdog(void)
 
 #ifdef CONFIG_PM
 
-#include <linux/device.h>
-
 static int nmi_pm_active; /* nmi_active before suspend */
 
 static int lapic_nmi_suspend(struct sys_device *dev, u32 state)

arch/x86_64/kernel/setup64.c

@@ -114,32 +114,37 @@ Valid options:
 
 __setup("noexec32=", nonx32_setup); 
 
-#ifndef  __GENERIC_PER_CPU
-
-unsigned long __per_cpu_offset[NR_CPUS];
-
+/*
+ * Great future plan:
+ * Declare PDA itself and support (irqstack,tss,pml4) as per cpu data.
+ * Always point %gs to its beginning
+ */
 void __init setup_per_cpu_areas(void)
 { 
-	unsigned long size, i;
-	unsigned char *ptr;
+	int i;
+	unsigned long size;
 
 	/* Copy section for each CPU (we discard the original) */
 	size = ALIGN(__per_cpu_end - __per_cpu_start, SMP_CACHE_BYTES);
-	if (!size)
-		return;
-
-	ptr = alloc_bootmem(size * NR_CPUS);
-
-	for (i = 0; i < NR_CPUS; i++, ptr += size) {
-		/* hide this from the compiler to avoid problems */ 
-		unsigned long offset;
-		asm("subq %[b],%0" : "=r" (offset) : "0" (ptr), [b] "r" (&__per_cpu_start));
-		__per_cpu_offset[i] = offset;
-		cpu_pda[i].cpudata_offset = offset;
-		memcpy(ptr, __per_cpu_start, size);
+#ifdef CONFIG_MODULES
+	if (size < PERCPU_ENOUGH_ROOM)
+		size = PERCPU_ENOUGH_ROOM;
+#endif
+
+	/* We don't support CPU hotplug, so only allocate as much as needed here */
+
+	int maxi = max_t(unsigned, numnodes, num_online_cpus());
+
+	for (i = 0; i < maxi; i++) {
+		/* If possible allocate on the node of the CPU.
+		   In case it doesn't exist round-robin nodes. */
+		unsigned char *ptr = alloc_bootmem_node(NODE_DATA(i % numnodes), 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);
 	}
 } 
-#endif
 
 void pda_init(int cpu)
 { 
@@ -153,7 +158,7 @@ void pda_init(int cpu)
 	pda->me = pda;
 	pda->cpunumber = cpu; 
 	pda->irqcount = -1;
-	pda->cpudata_offset = 0;
+	pda->data_offset = 0;
 	pda->kernelstack = 
 		(unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE; 
 	pda->active_mm = &init_mm;

arch/x86_64/mm/extable.c

@@ -5,6 +5,7 @@
 #include <linux/config.h>
 #include <linux/module.h>
 #include <linux/spinlock.h>
+#include <linux/init.h>
 #include <asm/uaccess.h>
 
 /* Simple binary search */
@@ -28,3 +29,29 @@ search_extable(const struct exception_table_entry *first,
         }
         return NULL;
 }
+
+/* When an exception handler is in an non standard section (like __init)
+   the fixup table can end up unordered. Fix that here. */
+static __init int check_extable(void)
+{
+	extern struct exception_table_entry __start___ex_table[];
+	extern struct exception_table_entry  __stop___ex_table[];
+	struct exception_table_entry *e;
+	int change;
+
+	/* The input is near completely presorted, which makes bubble sort the
+	   best (and simplest) sort algorithm. */
+	do {
+		change = 0;
+		for (e = __start___ex_table+1; e < __stop___ex_table; e++) {
+			if (e->insn < e[-1].insn) {
+				struct exception_table_entry tmp = e[-1];
+				e[-1] = e[0];
+				e[0] = tmp;
+				change = 1;
+			}
+		}
+	} while (change != 0);
+	return 0;
+}
+core_initcall(check_extable);

include/asm-x86_64/apic.h

@@ -84,10 +84,6 @@ extern int APIC_init_uniprocessor (void);
 extern void disable_APIC_timer(void);
 extern void enable_APIC_timer(void);
 
-#ifdef CONFIG_PM
-extern struct sys_device device_lapic;
-#endif
-
 extern int check_nmi_watchdog (void);
 
 extern unsigned int nmi_watchdog;

include/asm-x86_64/desc.h

@@ -7,6 +7,7 @@
 
 #ifndef __ASSEMBLY__
 
+#include <linux/string.h>
 #include <asm/segment.h>
 #include <asm/mmu.h>
 

include/asm-x86_64/hw_irq.h

@@ -164,7 +164,7 @@ static inline void x86_do_profile (struct pt_regs *regs)
 	atomic_inc((atomic_t *)&prof_buffer[rip]);
 }
 
-#ifdef CONFIG_SMP /*more of this file should probably be ifdefed SMP */
+#if defined(CONFIG_X86_IO_APIC) && defined(CONFIG_SMP)
 static inline void hw_resend_irq(struct hw_interrupt_type *h, unsigned int i) {
 	if (IO_APIC_IRQ(i))
 		send_IPI_self(IO_APIC_VECTOR(i));

include/asm-x86_64/local.h

+#ifndef _ARCH_X8664_LOCAL_H
+#define _ARCH_X8664_LOCAL_H
+
+#include <linux/percpu.h>
+
+typedef struct
+{
+	volatile unsigned int counter;
+} local_t;
+
+#define LOCAL_INIT(i)	{ (i) }
+
+#define local_read(v)	((v)->counter)
+#define local_set(v,i)	(((v)->counter) = (i))
+
+static __inline__ void local_inc(local_t *v)
+{
+	__asm__ __volatile__(
+		"incl %0"
+		:"=m" (v->counter)
+		:"m" (v->counter));
+}
+
+static __inline__ void local_dec(local_t *v)
+{
+	__asm__ __volatile__(
+		"decl %0"
+		:"=m" (v->counter)
+		:"m" (v->counter));
+}
+
+static __inline__ void local_add(unsigned long i, local_t *v)
+{
+	__asm__ __volatile__(
+		"addl %1,%0"
+		:"=m" (v->counter)
+		:"ir" (i), "m" (v->counter));
+}
+
+static __inline__ void local_sub(unsigned long i, local_t *v)
+{
+	__asm__ __volatile__(
+		"subl %1,%0"
+		:"=m" (v->counter)
+		:"ir" (i), "m" (v->counter));
+}
+
+/* On x86, these are no better than the atomic variants. */
+#define __local_inc(l)		local_inc(l)
+#define __local_dec(l)		local_dec(l)
+#define __local_add(i,l)	local_add((i),(l))
+#define __local_sub(i,l)	local_sub((i),(l))
+
+/* Use these for per-cpu local_t variables: on some archs they are
+ * much more efficient than these naive implementations.  Note they take
+ * a variable, not an address.
+ *
+ * This could be done better if we moved the per cpu data directly
+ * after GS.
+ */
+#define cpu_local_read(v)	local_read(&__get_cpu_var(v))
+#define cpu_local_set(v, i)	local_set(&__get_cpu_var(v), (i))
+#define cpu_local_inc(v)	local_inc(&__get_cpu_var(v))
+#define cpu_local_dec(v)	local_dec(&__get_cpu_var(v))
+#define cpu_local_add(i, v)	local_add((i), &__get_cpu_var(v))
+#define cpu_local_sub(i, v)	local_sub((i), &__get_cpu_var(v))
+
+#define __cpu_local_inc(v)	cpu_local_inc(v)
+#define __cpu_local_dec(v)	cpu_local_dec(v)
+#define __cpu_local_add(i, v)	cpu_local_add((i), (v))
+#define __cpu_local_sub(i, v)	cpu_local_sub((i), (v))
+
+#endif /* _ARCH_I386_LOCAL_H */

include/asm-x86_64/mpspec.h

@@ -164,11 +164,9 @@ enum mp_bustype {
 	MP_BUS_PCI,
 	MP_BUS_MCA
 };
-extern int mp_bus_id_to_type [MAX_MP_BUSSES];
-extern int mp_bus_id_to_node [MAX_MP_BUSSES];
-extern int mp_bus_id_to_local [MAX_MP_BUSSES];
-extern int quad_local_to_mp_bus_id [NR_CPUS/4][4];
+extern unsigned char mp_bus_id_to_type [MAX_MP_BUSSES];
 extern int mp_bus_id_to_pci_bus [MAX_MP_BUSSES];
+extern unsigned long mp_bus_to_cpumask [MAX_MP_BUSSES];
 
 extern unsigned int boot_cpu_physical_apicid;
 extern unsigned long phys_cpu_present_map;
@@ -177,11 +175,9 @@ extern void find_smp_config (void);
 extern void get_smp_config (void);
 extern int nr_ioapics;
 extern int apic_version [MAX_APICS];
-extern int mp_bus_id_to_type [MAX_MP_BUSSES];
 extern int mp_irq_entries;
 extern struct mpc_config_intsrc mp_irqs [MAX_IRQ_SOURCES];
 extern int mpc_default_type;
-extern int mp_bus_id_to_pci_bus [MAX_MP_BUSSES];
 extern int mp_current_pci_id;
 extern unsigned long mp_lapic_addr;
 extern int pic_mode;

include/asm-x86_64/nmi.h

@@ -48,4 +48,6 @@ static inline void unset_nmi_pm_callback(struct pm_dev * dev)
  
 extern void default_do_nmi(struct pt_regs *);
  
+extern void default_do_nmi(struct pt_regs *);
+
 #endif /* ASM_NMI_H */

include/asm-x86_64/pda.h

@@ -9,7 +9,7 @@
 /* Per processor datastructure. %gs points to it while the kernel runs */ 
 struct x8664_pda {
 	struct task_struct *pcurrent;	/* Current process */
-	unsigned long cpudata_offset;
+	unsigned long data_offset;	/* Per cpu data offset from linker address */
 	struct x8664_pda *me;	    /* Pointer to itself */  
 	unsigned long kernelstack;  /* TOS for current process */ 
 	unsigned long oldrsp; 	    /* user rsp for system call */

include/asm-x86_64/percpu.h

-#ifndef __ARCH_I386_PERCPU__
-#define __ARCH_I386_PERCPU__
+#ifndef _ASM_X8664_PERCPU_H_
+#define _ASM_X8664_PERCPU_H_
 
-#include <asm-generic/percpu.h>
+#include <asm/pda.h>
 
-#endif /* __ARCH_I386_PERCPU__ */
+#ifdef CONFIG_SMP
+
+/* Same as the generic code except that we cache the per cpu offset
+   in the pda. This gives an 3 instruction reference for per cpu data */
+
+#include <linux/compiler.h>
+#include <asm/pda.h>
+#define __my_cpu_offset() read_pda(data_offset)
+#define __per_cpu_offset(cpu) (cpu_pda[cpu].data_offset)
+
+/* Separate out the type, so (int[3], foo) works. */
+#define DEFINE_PER_CPU(type, name) \
+    __attribute__((__section__(".data.percpu"))) __typeof__(type) name##__per_cpu
+
+/* var is in discarded region: offset to particular copy we want */
+#define per_cpu(var, cpu) (*RELOC_HIDE(&var##__per_cpu, __per_cpu_offset(cpu)))
+#define __get_cpu_var(var) \
+	(*RELOC_HIDE(&var##__per_cpu, __my_cpu_offset()))
+
+static inline void percpu_modcopy(void *pcpudst, const void *src,
+				  unsigned long size)
+{
+	unsigned int i;
+	for (i = 0; i < NR_CPUS; i++)
+		if (cpu_possible(i))
+			memcpy(pcpudst + __per_cpu_offset(i), src, size);
+}
+
+extern void setup_per_cpu_areas(void);
+
+#else /* ! SMP */
+
+#define DEFINE_PER_CPU(type, name) \
+    __typeof__(type) name##__per_cpu
+
+#define per_cpu(var, cpu)			((void)cpu, var##__per_cpu)
+#define __get_cpu_var(var)			var##__per_cpu
+
+#endif	/* SMP */
+
+#define DECLARE_PER_CPU(type, name) extern __typeof__(type) name##__per_cpu
+
+#define EXPORT_PER_CPU_SYMBOL(var) EXPORT_SYMBOL(var##__per_cpu)
+#define EXPORT_PER_CPU_SYMBOL_GPL(var) EXPORT_SYMBOL_GPL(var##__per_cpu)
+
+DECLARE_PER_CPU(struct x8664_pda, per_cpu_pda);
+
+#endif

include/asm-x86_64/processor.h

@@ -181,6 +181,7 @@ static inline void clear_in_cr4 (unsigned long mask)
  * Size of io_bitmap in longwords: 32 is ports 0-0x3ff.
  */
 #define IO_BITMAP_SIZE	32
+#define IO_BITMAP_BYTES (IO_BITMAP_SIZE * 4)
 #define IO_BITMAP_OFFSET offsetof(struct tss_struct,io_bitmap)
 #define INVALID_IO_BITMAP_OFFSET 0x8000
 

include/asm-x86_64/smp.h

@@ -117,7 +117,7 @@ extern inline int safe_smp_processor_id(void)
 #ifndef CONFIG_SMP
 #define stack_smp_processor_id() 0
 #define safe_smp_processor_id() 0
-#define for_each_cpu(x) (x)=0;
+#define for_each_cpu(x,mask) (void)(mask), (x)=0;
 #define cpu_logical_map(x) (x)
 #else
 #include <asm/thread_info.h>

include/asm-x86_64/topology.h

@@ -5,6 +5,8 @@
 
 #ifdef CONFIG_DISCONTIGMEM
 
+#include <asm/mpspec.h>
+
 /* Map the K8 CPU local memory controllers to a simple 1:1 CPU:NODE topology */
 
 extern int fake_node;
@@ -17,6 +19,11 @@ extern unsigned long cpu_online_map;
 #define node_to_cpu_mask(node)	(fake_node ? cpu_online_map : (1UL << (node)))
 #define node_to_memblk(node)		(node)
 
+static inline unsigned long pcibus_to_cpumask(int bus)
+{
+	return mp_bus_to_cpumask[bus] & cpu_online_map;
+}
+
 #define NODE_BALANCE_RATE 30	/* CHECKME */ 
 
 #endif

include/asm-x86_64/unistd.h

@@ -526,8 +526,12 @@ __SYSCALL(__NR_exit_group, sys_exit_group)
 __SYSCALL(__NR_epoll_wait, sys_epoll_wait)
 #define __NR_epoll_ctl		233
 __SYSCALL(__NR_epoll_ctl, sys_epoll_ctl)
+#define __NR_tgkill		234
+__SYSCALL(__NR_tgkill, sys_tgkill)
+#define __NR_utimes		235
+__SYSCALL(__NR_utimes, sys_utimes)
 
-#define __NR_syscall_max __NR_epoll_ctl
+#define __NR_syscall_max __NR_utimes
 #ifndef __NO_STUBS
 
 /* user-visible error numbers are in the range -1 - -4095 */