x86_64: prepare shared kernel/process.c
authorThomas Gleixner <tglx@linutronix.de>
Thu, 11 Oct 2007 09:15:10 +0000 (11:15 +0200)
committerThomas Gleixner <tglx@linutronix.de>
Thu, 11 Oct 2007 09:15:10 +0000 (11:15 +0200)
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
arch/x86_64/kernel/Makefile
arch/x86_64/kernel/process.c [deleted file]
arch/x86_64/kernel/process_64.c [new file with mode: 0644]

index d50ba296437c45d200e8ee7aa621f578948a222c..976ea5e7222bdbde3a86eac3e66ce52c0ef68dd0 100644 (file)
@@ -4,7 +4,7 @@
 
 extra-y        := head.o head64.o init_task_64.o vmlinux.lds
 EXTRA_AFLAGS   := -traditional
-obj-y  := process.o signal_64.o entry.o traps_64.o irq_64.o \
+obj-y  := process_64.o signal_64.o entry.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.o vsyscall_64.o \
                setup64.o bootflag.o e820_64.o reboot_64.o quirks.o i8237.o \
diff --git a/arch/x86_64/kernel/process.c b/arch/x86_64/kernel/process.c
deleted file mode 100644 (file)
index 9895655..0000000
+++ /dev/null
@@ -1,903 +0,0 @@
-/*
- *  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;
-}
diff --git a/arch/x86_64/kernel/process_64.c b/arch/x86_64/kernel/process_64.c
new file mode 100644 (file)
index 0000000..9895655
--- /dev/null
@@ -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;
+}