Merge git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6
[safe/jmp/linux-2.6] / arch / x86 / kernel / traps_64.c
index df690c3..7a31f10 100644 (file)
  * 'Traps.c' handles hardware traps and faults after we have saved some
  * state in 'entry.S'.
  */
-#include <linux/sched.h>
+#include <linux/moduleparam.h>
+#include <linux/interrupt.h>
+#include <linux/kallsyms.h>
+#include <linux/spinlock.h>
+#include <linux/kprobes.h>
+#include <linux/uaccess.h>
+#include <linux/utsname.h>
+#include <linux/kdebug.h>
 #include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/ptrace.h>
 #include <linux/string.h>
+#include <linux/unwind.h>
+#include <linux/delay.h>
 #include <linux/errno.h>
-#include <linux/ptrace.h>
+#include <linux/kexec.h>
+#include <linux/sched.h>
 #include <linux/timer.h>
-#include <linux/mm.h>
 #include <linux/init.h>
-#include <linux/delay.h>
-#include <linux/spinlock.h>
-#include <linux/interrupt.h>
-#include <linux/kallsyms.h>
-#include <linux/module.h>
-#include <linux/moduleparam.h>
-#include <linux/nmi.h>
-#include <linux/kprobes.h>
-#include <linux/kexec.h>
-#include <linux/unwind.h>
-#include <linux/uaccess.h>
 #include <linux/bug.h>
-#include <linux/kdebug.h>
+#include <linux/nmi.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/io.h>
 
 #if defined(CONFIG_EDAC)
 #include <linux/edac.h>
 #endif
 
-#include <asm/system.h>
-#include <asm/io.h>
-#include <asm/atomic.h>
+#include <asm/stacktrace.h>
+#include <asm/processor.h>
 #include <asm/debugreg.h>
+#include <asm/atomic.h>
+#include <asm/system.h>
+#include <asm/unwind.h>
 #include <asm/desc.h>
 #include <asm/i387.h>
-#include <asm/processor.h>
-#include <asm/unwind.h>
-#include <asm/smp.h>
 #include <asm/pgalloc.h>
-#include <asm/pda.h>
 #include <asm/proto.h>
-#include <asm/nmi.h>
-#include <asm/stacktrace.h>
+#include <asm/pda.h>
+#include <asm/traps.h>
+
+#include <mach_traps.h>
 
-asmlinkage void divide_error(void);
-asmlinkage void debug(void);
-asmlinkage void nmi(void);
-asmlinkage void int3(void);
-asmlinkage void overflow(void);
-asmlinkage void bounds(void);
-asmlinkage void invalid_op(void);
-asmlinkage void device_not_available(void);
-asmlinkage void double_fault(void);
-asmlinkage void coprocessor_segment_overrun(void);
-asmlinkage void invalid_TSS(void);
-asmlinkage void segment_not_present(void);
-asmlinkage void stack_segment(void);
-asmlinkage void general_protection(void);
-asmlinkage void page_fault(void);
-asmlinkage void coprocessor_error(void);
-asmlinkage void simd_coprocessor_error(void);
-asmlinkage void reserved(void);
-asmlinkage void alignment_check(void);
-asmlinkage void machine_check(void);
-asmlinkage void spurious_interrupt_bug(void);
+int panic_on_unrecovered_nmi;
+int kstack_depth_to_print = 12;
+static unsigned int code_bytes = 64;
+static int ignore_nmis;
+static int die_counter;
 
 static inline void conditional_sti(struct pt_regs *regs)
 {
-       if (regs->eflags & X86_EFLAGS_IF)
+       if (regs->flags & X86_EFLAGS_IF)
                local_irq_enable();
 }
 
 static inline void preempt_conditional_sti(struct pt_regs *regs)
 {
-       preempt_disable();
-       if (regs->eflags & X86_EFLAGS_IF)
+       inc_preempt_count();
+       if (regs->flags & X86_EFLAGS_IF)
                local_irq_enable();
 }
 
 static inline void preempt_conditional_cli(struct pt_regs *regs)
 {
-       if (regs->eflags & X86_EFLAGS_IF)
+       if (regs->flags & X86_EFLAGS_IF)
                local_irq_disable();
        /* Make sure to not schedule here because we could be running
           on an exception stack. */
-       preempt_enable_no_resched();
+       dec_preempt_count();
 }
 
-int kstack_depth_to_print = 12;
-
-#ifdef CONFIG_KALLSYMS
-void printk_address(unsigned long address)
+void printk_address(unsigned long address, int reliable)
 {
-       unsigned long offset = 0, symsize;
-       const char *symname;
-       char *modname;
-       char *delim = ":";
-       char namebuf[128];
-
-       symname = kallsyms_lookup(address, &symsize, &offset,
-                                       &modname, namebuf);
-       if (!symname) {
-               printk(" [<%016lx>]\n", address);
-               return;
-       }
-       if (!modname)
-               modname = delim = "";           
-       printk(" [<%016lx>] %s%s%s%s+0x%lx/0x%lx\n",
-               address, delim, modname, delim, symname, offset, symsize);
+       printk(" [<%016lx>] %s%pS\n",
+                       address, reliable ?     "" : "? ", (void *) address);
 }
-#else
-void printk_address(unsigned long address)
-{
-       printk(" [<%016lx>]\n", address);
-}
-#endif
 
 static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
                                        unsigned *usedp, char **idp)
@@ -134,7 +98,8 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
                [STACKFAULT_STACK - 1] = "#SS",
                [MCE_STACK - 1] = "#MC",
 #if DEBUG_STKSZ > EXCEPTION_STKSZ
-               [N_EXCEPTION_STACKS ... N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
+               [N_EXCEPTION_STACKS ...
+                       N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
 #endif
        };
        unsigned k;
@@ -198,68 +163,95 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
        return NULL;
 }
 
-#define MSG(txt) ops->warning(data, txt)
-
 /*
- * x86-64 can have upto three kernel stacks: 
+ * x86-64 can have up to three kernel stacks:
  * process stack
  * interrupt stack
  * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
  */
 
-static inline int valid_stack_ptr(struct thread_info *tinfo, void *p)
+static inline int valid_stack_ptr(struct thread_info *tinfo,
+                       void *p, unsigned int size, void *end)
+{
+       void *t = tinfo;
+       if (end) {
+               if (p < end && p >= (end-THREAD_SIZE))
+                       return 1;
+               else
+                       return 0;
+       }
+       return p > t && p < t + THREAD_SIZE - size;
+}
+
+/* The form of the top of the frame on the stack */
+struct stack_frame {
+       struct stack_frame *next_frame;
+       unsigned long return_address;
+};
+
+static inline unsigned long
+print_context_stack(struct thread_info *tinfo,
+               unsigned long *stack, unsigned long bp,
+               const struct stacktrace_ops *ops, void *data,
+               unsigned long *end)
 {
-       void *t = (void *)tinfo;
-        return p > t && p < t + THREAD_SIZE - 3;
+       struct stack_frame *frame = (struct stack_frame *)bp;
+
+       while (valid_stack_ptr(tinfo, stack, sizeof(*stack), end)) {
+               unsigned long addr;
+
+               addr = *stack;
+               if (__kernel_text_address(addr)) {
+                       if ((unsigned long) stack == bp + 8) {
+                               ops->address(data, addr, 1);
+                               frame = frame->next_frame;
+                               bp = (unsigned long) frame;
+                       } else {
+                               ops->address(data, addr, bp == 0);
+                       }
+               }
+               stack++;
+       }
+       return bp;
 }
 
-void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
-               unsigned long *stack,
+void dump_trace(struct task_struct *task, struct pt_regs *regs,
+               unsigned long *stack, unsigned long bp,
                const struct stacktrace_ops *ops, void *data)
 {
        const unsigned cpu = get_cpu();
-       unsigned long *irqstack_end = (unsigned long*)cpu_pda(cpu)->irqstackptr;
+       unsigned long *irqstack_end = (unsigned long *)cpu_pda(cpu)->irqstackptr;
        unsigned used = 0;
        struct thread_info *tinfo;
 
-       if (!tsk)
-               tsk = current;
+       if (!task)
+               task = current;
 
        if (!stack) {
                unsigned long dummy;
                stack = &dummy;
-               if (tsk && tsk != current)
-                       stack = (unsigned long *)tsk->thread.rsp;
+               if (task && task != current)
+                       stack = (unsigned long *)task->thread.sp;
        }
 
-       /*
-        * Print function call entries within a stack. 'cond' is the
-        * "end of stackframe" condition, that the 'stack++'
-        * iteration will eventually trigger.
-        */
-#define HANDLE_STACK(cond) \
-       do while (cond) { \
-               unsigned long addr = *stack++; \
-               /* Use unlocked access here because except for NMIs     \
-                  we should be already protected against module unloads */ \
-               if (__kernel_text_address(addr)) { \
-                       /* \
-                        * If the address is either in the text segment of the \
-                        * kernel, or in the region which contains vmalloc'ed \
-                        * memory, it *may* be the address of a calling \
-                        * routine; if so, print it so that someone tracing \
-                        * down the cause of the crash will be able to figure \
-                        * out the call path that was taken. \
-                        */ \
-                       ops->address(data, addr);   \
-               } \
-       } while (0)
+#ifdef CONFIG_FRAME_POINTER
+       if (!bp) {
+               if (task == current) {
+                       /* Grab bp right from our regs */
+                       asm("movq %%rbp, %0" : "=r" (bp) : );
+               } else {
+                       /* bp is the last reg pushed by switch_to */
+                       bp = *(unsigned long *) task->thread.sp;
+               }
+       }
+#endif
 
        /*
         * Print function call entries in all stacks, starting at the
         * current stack address. If the stacks consist of nested
         * exceptions
         */
+       tinfo = task_thread_info(task);
        for (;;) {
                char *id;
                unsigned long *estack_end;
@@ -269,7 +261,9 @@ void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
                if (estack_end) {
                        if (ops->stack(data, id) < 0)
                                break;
-                       HANDLE_STACK (stack < estack_end);
+
+                       bp = print_context_stack(tinfo, stack, bp, ops,
+                                                       data, estack_end);
                        ops->stack(data, "<EOE>");
                        /*
                         * We link to the next stack via the
@@ -287,7 +281,8 @@ void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
                        if (stack >= irqstack && stack < irqstack_end) {
                                if (ops->stack(data, "IRQ") < 0)
                                        break;
-                               HANDLE_STACK (stack < irqstack_end);
+                               bp = print_context_stack(tinfo, stack, bp,
+                                               ops, data, irqstack_end);
                                /*
                                 * We link to the next stack (which would be
                                 * the process stack normally) the last
@@ -305,9 +300,7 @@ void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
        /*
         * This handles the process stack:
         */
-       tinfo = task_thread_info(tsk);
-       HANDLE_STACK (valid_stack_ptr(tinfo, stack));
-#undef HANDLE_STACK
+       bp = print_context_stack(tinfo, stack, bp, ops, data, NULL);
        put_cpu();
 }
 EXPORT_SYMBOL(dump_trace);
@@ -330,10 +323,10 @@ static int print_trace_stack(void *data, char *name)
        return 0;
 }
 
-static void print_trace_address(void *data, unsigned long addr)
+static void print_trace_address(void *data, unsigned long addr, int reliable)
 {
        touch_nmi_watchdog();
-       printk_address(addr);
+       printk_address(addr, reliable);
 }
 
 static const struct stacktrace_ops print_trace_ops = {
@@ -343,35 +336,46 @@ static const struct stacktrace_ops print_trace_ops = {
        .address = print_trace_address,
 };
 
-void
-show_trace(struct task_struct *tsk, struct pt_regs *regs, unsigned long *stack)
+static void
+show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
+               unsigned long *stack, unsigned long bp, char *log_lvl)
 {
-       printk("\nCall Trace:\n");
-       dump_trace(tsk, regs, stack, &print_trace_ops, NULL);
-       printk("\n");
+       printk("Call Trace:\n");
+       dump_trace(task, regs, stack, bp, &print_trace_ops, log_lvl);
+}
+
+void show_trace(struct task_struct *task, struct pt_regs *regs,
+               unsigned long *stack, unsigned long bp)
+{
+       show_trace_log_lvl(task, regs, stack, bp, "");
 }
 
 static void
-_show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long *rsp)
+show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
+               unsigned long *sp, unsigned long bp, char *log_lvl)
 {
        unsigned long *stack;
        int i;
        const int cpu = smp_processor_id();
-       unsigned long *irqstack_end = (unsigned long *) (cpu_pda(cpu)->irqstackptr);
-       unsigned long *irqstack = (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
+       unsigned long *irqstack_end =
+               (unsigned long *) (cpu_pda(cpu)->irqstackptr);
+       unsigned long *irqstack =
+               (unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
 
-       // debugging aid: "show_stack(NULL, NULL);" prints the
-       // back trace for this cpu.
+       /*
+        * debugging aid: "show_stack(NULL, NULL);" prints the
+        * back trace for this cpu.
+        */
 
-       if (rsp == NULL) {
-               if (tsk)
-                       rsp = (unsigned long *)tsk->thread.rsp;
+       if (sp == NULL) {
+               if (task)
+                       sp = (unsigned long *)task->thread.sp;
                else
-                       rsp = (unsigned long *)&rsp;
+                       sp = (unsigned long *)&sp;
        }
 
-       stack = rsp;
-       for(i=0; i < kstack_depth_to_print; i++) {
+       stack = sp;
+       for (i = 0; i < kstack_depth_to_print; i++) {
                if (stack >= irqstack && stack <= irqstack_end) {
                        if (stack == irqstack_end) {
                                stack = (unsigned long *) (irqstack_end[-1]);
@@ -386,12 +390,13 @@ _show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long *rsp)
                printk(" %016lx", *stack++);
                touch_nmi_watchdog();
        }
-       show_trace(tsk, regs, rsp);
+       printk("\n");
+       show_trace_log_lvl(task, regs, sp, bp, log_lvl);
 }
 
-void show_stack(struct task_struct *tsk, unsigned long * rsp)
+void show_stack(struct task_struct *task, unsigned long *sp)
 {
-       _show_stack(tsk, NULL, rsp);
+       show_stack_log_lvl(task, NULL, sp, 0, "");
 }
 
 /*
@@ -399,21 +404,31 @@ void show_stack(struct task_struct *tsk, unsigned long * rsp)
  */
 void dump_stack(void)
 {
-       unsigned long dummy;
-       show_trace(NULL, NULL, &dummy);
-}
+       unsigned long bp = 0;
+       unsigned long stack;
 
+#ifdef CONFIG_FRAME_POINTER
+       if (!bp)
+               asm("movq %%rbp, %0" : "=r" (bp) : );
+#endif
+
+       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);
+       show_trace(NULL, NULL, &stack, bp);
+}
 EXPORT_SYMBOL(dump_stack);
 
 void show_registers(struct pt_regs *regs)
 {
        int i;
-       int in_kernel = !user_mode(regs);
-       unsigned long rsp;
+       unsigned long sp;
        const int cpu = smp_processor_id();
        struct task_struct *cur = cpu_pda(cpu)->pcurrent;
 
-       rsp = regs->rsp;
+       sp = regs->sp;
        printk("CPU %d ", cpu);
        __show_regs(regs);
        printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
@@ -423,45 +438,49 @@ void show_registers(struct pt_regs *regs)
         * When in-kernel, we also print out the stack and code at the
         * time of the fault..
         */
-       if (in_kernel) {
+       if (!user_mode(regs)) {
+               unsigned int code_prologue = code_bytes * 43 / 64;
+               unsigned int code_len = code_bytes;
+               unsigned char c;
+               u8 *ip;
+
                printk("Stack: ");
-               _show_stack(NULL, regs, (unsigned long*)rsp);
+               show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
+                               regs->bp, "");
 
-               printk("\nCode: ");
-               if (regs->rip < PAGE_OFFSET)
-                       goto bad;
+               printk(KERN_EMERG "Code: ");
 
-               for (i=0; i<20; i++) {
-                       unsigned char c;
-                       if (__get_user(c, &((unsigned char*)regs->rip)[i])) {
-bad:
+               ip = (u8 *)regs->ip - code_prologue;
+               if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) {
+                       /* try starting at RIP */
+                       ip = (u8 *)regs->ip;
+                       code_len = code_len - code_prologue + 1;
+               }
+               for (i = 0; i < code_len; i++, ip++) {
+                       if (ip < (u8 *)PAGE_OFFSET ||
+                                       probe_kernel_address(ip, c)) {
                                printk(" Bad RIP value.");
                                break;
                        }
-                       printk("%02x ", c);
+                       if (ip == (u8 *)regs->ip)
+                               printk("<%02x> ", c);
+                       else
+                               printk("%02x ", c);
                }
        }
        printk("\n");
-}      
+}
 
-int is_valid_bugaddr(unsigned long rip)
+int is_valid_bugaddr(unsigned long ip)
 {
        unsigned short ud2;
 
-       if (__copy_from_user(&ud2, (const void __user *) rip, sizeof(ud2)))
+       if (__copy_from_user(&ud2, (const void __user *) ip, sizeof(ud2)))
                return 0;
 
        return ud2 == 0x0b0f;
 }
 
-#ifdef CONFIG_BUG
-void out_of_line_bug(void)
-{ 
-       BUG(); 
-} 
-EXPORT_SYMBOL(out_of_line_bug);
-#endif
-
 static raw_spinlock_t die_lock = __RAW_SPIN_LOCK_UNLOCKED;
 static int die_owner = -1;
 static unsigned int die_nest_count;
@@ -477,7 +496,7 @@ unsigned __kprobes long oops_begin(void)
        raw_local_irq_save(flags);
        cpu = smp_processor_id();
        if (!__raw_spin_trylock(&die_lock)) {
-               if (cpu == die_owner) 
+               if (cpu == die_owner)
                        /* nested oops. should stop eventually */;
                else
                        __raw_spin_lock(&die_lock);
@@ -489,8 +508,8 @@ unsigned __kprobes long oops_begin(void)
        return flags;
 }
 
-void __kprobes oops_end(unsigned long flags)
-{ 
+void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr)
+{
        die_owner = -1;
        bust_spinlocks(0);
        die_nest_count--;
@@ -498,15 +517,19 @@ void __kprobes oops_end(unsigned long flags)
                /* Nest count reaches zero, release the lock. */
                __raw_spin_unlock(&die_lock);
        raw_local_irq_restore(flags);
+       if (!regs) {
+               oops_exit();
+               return;
+       }
        if (panic_on_oops)
                panic("Fatal exception");
        oops_exit();
+       do_exit(signr);
 }
 
-void __kprobes __die(const char * str, struct pt_regs * regs, long err)
+int __kprobes __die(const char *str, struct pt_regs *regs, long err)
 {
-       static int die_counter;
-       printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff,++die_counter);
+       printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff, ++die_counter);
 #ifdef CONFIG_PREEMPT
        printk("PREEMPT ");
 #endif
@@ -517,134 +540,140 @@ void __kprobes __die(const char * str, struct pt_regs * regs, long err)
        printk("DEBUG_PAGEALLOC");
 #endif
        printk("\n");
-       notify_die(DIE_OOPS, str, regs, err, current->thread.trap_no, SIGSEGV);
+       if (notify_die(DIE_OOPS, str, regs, err,
+                       current->thread.trap_no, SIGSEGV) == NOTIFY_STOP)
+               return 1;
+
        show_registers(regs);
        add_taint(TAINT_DIE);
        /* Executive summary in case the oops scrolled away */
        printk(KERN_ALERT "RIP ");
-       printk_address(regs->rip); 
-       printk(" RSP <%016lx>\n", regs->rsp); 
+       printk_address(regs->ip, 1);
+       printk(" RSP <%016lx>\n", regs->sp);
        if (kexec_should_crash(current))
                crash_kexec(regs);
+       return 0;
 }
 
-void die(const char * str, struct pt_regs * regs, long err)
+void die(const char *str, struct pt_regs *regs, long err)
 {
        unsigned long flags = oops_begin();
 
        if (!user_mode(regs))
-               report_bug(regs->rip, regs);
+               report_bug(regs->ip, regs);
 
-       __die(str, regs, err);
-       oops_end(flags);
-       do_exit(SIGSEGV); 
+       if (__die(str, regs, err))
+               regs = NULL;
+       oops_end(flags, regs, SIGSEGV);
 }
 
-void __kprobes die_nmi(char *str, struct pt_regs *regs, int do_panic)
+notrace __kprobes void
+die_nmi(char *str, struct pt_regs *regs, int do_panic)
 {
-       unsigned long flags = oops_begin();
+       unsigned long flags;
 
+       if (notify_die(DIE_NMIWATCHDOG, str, regs, 0, 2, SIGINT) == NOTIFY_STOP)
+               return;
+
+       flags = oops_begin();
        /*
         * We are in trouble anyway, lets at least try
         * to get a message out.
         */
-       printk(str, smp_processor_id());
+       printk(KERN_EMERG "%s", str);
+       printk(" on CPU%d, ip %08lx, registers:\n",
+               smp_processor_id(), regs->ip);
        show_registers(regs);
        if (kexec_should_crash(current))
                crash_kexec(regs);
        if (do_panic || panic_on_oops)
                panic("Non maskable interrupt");
-       oops_end(flags);
+       oops_end(flags, NULL, SIGBUS);
        nmi_exit();
        local_irq_enable();
-       do_exit(SIGSEGV);
+       do_exit(SIGBUS);
 }
 
-static void __kprobes do_trap(int trapnr, int signr, char *str,
-                             struct pt_regs * regs, long error_code,
-                             siginfo_t *info)
+static void __kprobes
+do_trap(int trapnr, int signr, char *str, struct pt_regs *regs,
+       long error_code, siginfo_t *info)
 {
        struct task_struct *tsk = current;
 
-       if (user_mode(regs)) {
-               /*
-                * We want error_code and trap_no set for userspace
-                * faults and kernelspace faults which result in
-                * die(), but not kernelspace faults which are fixed
-                * up.  die() gives the process no chance to handle
-                * the signal and notice the kernel fault information,
-                * so that won't result in polluting the information
-                * about previously queued, but not yet delivered,
-                * faults.  See also do_general_protection below.
-                */
-               tsk->thread.error_code = error_code;
-               tsk->thread.trap_no = trapnr;
-
-               if (show_unhandled_signals && unhandled_signal(tsk, signr) &&
-                   printk_ratelimit())
-                       printk(KERN_INFO
-                              "%s[%d] trap %s rip:%lx rsp:%lx error:%lx\n",
-                              tsk->comm, tsk->pid, str,
-                              regs->rip, regs->rsp, error_code); 
+       if (!user_mode(regs))
+               goto kernel_trap;
 
-               if (info)
-                       force_sig_info(signr, info, tsk);
-               else
-                       force_sig(signr, tsk);
-               return;
+       /*
+        * We want error_code and trap_no set for userspace faults and
+        * kernelspace faults which result in die(), but not
+        * kernelspace faults which are fixed up.  die() gives the
+        * process no chance to handle the signal and notice the
+        * kernel fault information, so that won't result in polluting
+        * the information about previously queued, but not yet
+        * delivered, faults.  See also do_general_protection below.
+        */
+       tsk->thread.error_code = error_code;
+       tsk->thread.trap_no = trapnr;
+
+       if (show_unhandled_signals && unhandled_signal(tsk, signr) &&
+           printk_ratelimit()) {
+               printk(KERN_INFO
+                      "%s[%d] trap %s ip:%lx sp:%lx error:%lx",
+                      tsk->comm, tsk->pid, str,
+                      regs->ip, regs->sp, error_code);
+               print_vma_addr(" in ", regs->ip);
+               printk("\n");
        }
 
+       if (info)
+               force_sig_info(signr, info, tsk);
+       else
+               force_sig(signr, tsk);
+       return;
 
-       /* kernel trap */ 
-       {            
-               const struct exception_table_entry *fixup;
-               fixup = search_exception_tables(regs->rip);
-               if (fixup)
-                       regs->rip = fixup->fixup;
-               else {
-                       tsk->thread.error_code = error_code;
-                       tsk->thread.trap_no = trapnr;
-                       die(str, regs, error_code);
-               }
-               return;
+kernel_trap:
+       if (!fixup_exception(regs)) {
+               tsk->thread.error_code = error_code;
+               tsk->thread.trap_no = trapnr;
+               die(str, regs, error_code);
        }
+       return;
 }
 
 #define DO_ERROR(trapnr, signr, str, name) \
-asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
-{ \
-       if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-                                                       == NOTIFY_STOP) \
-               return; \
+asmlinkage void do_##name(struct pt_regs *regs, long error_code)       \
+{                                                                      \
+       if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr)  \
+                                                       == NOTIFY_STOP) \
+               return;                                                 \
        conditional_sti(regs);                                          \
-       do_trap(trapnr, signr, str, regs, error_code, NULL); \
+       do_trap(trapnr, signr, str, regs, error_code, NULL);            \
 }
 
-#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
-asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
-{ \
-       siginfo_t info; \
-       info.si_signo = signr; \
-       info.si_errno = 0; \
-       info.si_code = sicode; \
-       info.si_addr = (void __user *)siaddr; \
-       if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
-                                                       == NOTIFY_STOP) \
-               return; \
+#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr)                \
+asmlinkage void do_##name(struct pt_regs *regs, long error_code)       \
+{                                                                      \
+       siginfo_t info;                                                 \
+       info.si_signo = signr;                                          \
+       info.si_errno = 0;                                              \
+       info.si_code = sicode;                                          \
+       info.si_addr = (void __user *)siaddr;                           \
+       trace_hardirqs_fixup();                                         \
+       if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr)  \
+                                                       == NOTIFY_STOP) \
+               return;                                                 \
        conditional_sti(regs);                                          \
-       do_trap(trapnr, signr, str, regs, error_code, &info); \
+       do_trap(trapnr, signr, str, regs, error_code, &info);           \
 }
 
-DO_ERROR_INFO( 0, SIGFPE,  "divide error", divide_error, FPE_INTDIV, regs->rip)
-DO_ERROR( 4, SIGSEGV, "overflow", overflow)
-DO_ERROR( 5, SIGSEGV, "bounds", bounds)
-DO_ERROR_INFO( 6, SIGILL,  "invalid opcode", invalid_op, ILL_ILLOPN, regs->rip)
-DO_ERROR( 7, SIGSEGV, "device not available", device_not_available)
-DO_ERROR( 9, SIGFPE,  "coprocessor segment overrun", coprocessor_segment_overrun)
+DO_ERROR_INFO(0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip)
+DO_ERROR(4, SIGSEGV, "overflow", overflow)
+DO_ERROR(5, SIGSEGV, "bounds", bounds)
+DO_ERROR_INFO(6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip)
+DO_ERROR(9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun)
 DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
-DO_ERROR(11, SIGBUS,  "segment not present", segment_not_present)
+DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
 DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
-DO_ERROR(18, SIGSEGV, "reserved", reserved)
 
 /* Runs on IST stack */
 asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code)
@@ -657,7 +686,7 @@ asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code)
        preempt_conditional_cli(regs);
 }
 
-asmlinkage void do_double_fault(struct pt_regs * regs, long error_code)
+asmlinkage void do_double_fault(struct pt_regs *regs, long error_code)
 {
        static const char str[] = "double fault";
        struct task_struct *tsk = current;
@@ -674,55 +703,54 @@ asmlinkage void do_double_fault(struct pt_regs * regs, long error_code)
                die(str, regs, error_code);
 }
 
-asmlinkage void __kprobes do_general_protection(struct pt_regs * regs,
-                                               long error_code)
+asmlinkage void __kprobes
+do_general_protection(struct pt_regs *regs, long error_code)
 {
-       struct task_struct *tsk = current;
+       struct task_struct *tsk;
 
        conditional_sti(regs);
 
-       if (user_mode(regs)) {
-               tsk->thread.error_code = error_code;
-               tsk->thread.trap_no = 13;
+       tsk = current;
+       if (!user_mode(regs))
+               goto gp_in_kernel;
 
-               if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) &&
-                   printk_ratelimit())
-                       printk(KERN_INFO
-                      "%s[%d] general protection rip:%lx rsp:%lx error:%lx\n",
-                              tsk->comm, tsk->pid,
-                              regs->rip, regs->rsp, error_code); 
+       tsk->thread.error_code = error_code;
+       tsk->thread.trap_no = 13;
+
+       if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) &&
+                       printk_ratelimit()) {
+               printk(KERN_INFO
+                       "%s[%d] general protection ip:%lx sp:%lx error:%lx",
+                       tsk->comm, tsk->pid,
+                       regs->ip, regs->sp, error_code);
+               print_vma_addr(" in ", regs->ip);
+               printk("\n");
+       }
 
-               force_sig(SIGSEGV, tsk);
+       force_sig(SIGSEGV, tsk);
+       return;
+
+gp_in_kernel:
+       if (fixup_exception(regs))
                return;
-       } 
-
-       /* kernel gp */
-       {
-               const struct exception_table_entry *fixup;
-               fixup = search_exception_tables(regs->rip);
-               if (fixup) {
-                       regs->rip = fixup->fixup;
-                       return;
-               }
 
-               tsk->thread.error_code = error_code;
-               tsk->thread.trap_no = 13;
-               if (notify_die(DIE_GPF, "general protection fault", regs,
-                                       error_code, 13, SIGSEGV) == NOTIFY_STOP)
-                       return;
-               die("general protection fault", regs, error_code);
-       }
+       tsk->thread.error_code = error_code;
+       tsk->thread.trap_no = 13;
+       if (notify_die(DIE_GPF, "general protection fault", regs,
+                               error_code, 13, SIGSEGV) == NOTIFY_STOP)
+               return;
+       die("general protection fault", regs, error_code);
 }
 
-static __kprobes void
-mem_parity_error(unsigned char reason, struct pt_regs * regs)
+static notrace __kprobes void
+mem_parity_error(unsigned char reason, struct pt_regs *regs)
 {
        printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
                reason);
        printk(KERN_EMERG "You have some hardware problem, likely on the PCI bus.\n");
 
 #if defined(CONFIG_EDAC)
-       if(edac_handler_set()) {
+       if (edac_handler_set()) {
                edac_atomic_assert_error();
                return;
        }
@@ -738,8 +766,8 @@ mem_parity_error(unsigned char reason, struct pt_regs * regs)
        outb(reason, 0x61);
 }
 
-static __kprobes void
-io_check_error(unsigned char reason, struct pt_regs * regs)
+static notrace __kprobes void
+io_check_error(unsigned char reason, struct pt_regs *regs)
 {
        printk("NMI: IOCK error (debug interrupt?)\n");
        show_registers(regs);
@@ -752,9 +780,12 @@ io_check_error(unsigned char reason, struct pt_regs * regs)
        outb(reason, 0x61);
 }
 
-static __kprobes void
-unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
+static notrace __kprobes void
+unknown_nmi_error(unsigned char reason, struct pt_regs *regs)
 {
+       if (notify_die(DIE_NMIUNKNOWN, "nmi", regs, reason, 2, SIGINT) ==
+                       NOTIFY_STOP)
+               return;
        printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
                reason);
        printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n");
@@ -767,14 +798,14 @@ unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
 
 /* Runs on IST stack. This code must keep interrupts off all the time.
    Nested NMIs are prevented by the CPU. */
-asmlinkage __kprobes void default_do_nmi(struct pt_regs *regs)
+asmlinkage notrace __kprobes void default_do_nmi(struct pt_regs *regs)
 {
        unsigned char reason = 0;
        int cpu;
 
        cpu = smp_processor_id();
 
-       /* Only the BSP gets external NMIs from the system.  */
+       /* Only the BSP gets external NMIs from the system. */
        if (!cpu)
                reason = get_nmi_reason();
 
@@ -786,30 +817,57 @@ asmlinkage __kprobes void default_do_nmi(struct pt_regs *regs)
                 * Ok, so this is none of the documented NMI sources,
                 * so it must be the NMI watchdog.
                 */
-               if (nmi_watchdog_tick(regs,reason))
+               if (nmi_watchdog_tick(regs, reason))
                        return;
-               if (!do_nmi_callback(regs,cpu))
+               if (!do_nmi_callback(regs, cpu))
                        unknown_nmi_error(reason, regs);
 
                return;
        }
        if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
-               return; 
+               return;
 
        /* AK: following checks seem to be broken on modern chipsets. FIXME */
-
        if (reason & 0x80)
                mem_parity_error(reason, regs);
        if (reason & 0x40)
                io_check_error(reason, regs);
 }
 
+asmlinkage notrace __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();
+}
+
+void stop_nmi(void)
+{
+       acpi_nmi_disable();
+       ignore_nmis++;
+}
+
+void restart_nmi(void)
+{
+       ignore_nmis--;
+       acpi_nmi_enable();
+}
+
 /* runs on IST stack. */
-asmlinkage void __kprobes do_int3(struct pt_regs * regs, long error_code)
+asmlinkage void __kprobes do_int3(struct pt_regs *regs, long error_code)
 {
-       if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) {
+       trace_hardirqs_fixup();
+
+       if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP)
+                       == NOTIFY_STOP)
                return;
-       }
+
        preempt_conditional_sti(regs);
        do_trap(3, SIGTRAP, "int3", regs, error_code, NULL);
        preempt_conditional_cli(regs);
@@ -822,30 +880,38 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
 {
        struct pt_regs *regs = eregs;
        /* Did already sync */
-       if (eregs == (struct pt_regs *)eregs->rsp)
+       if (eregs == (struct pt_regs *)eregs->sp)
                ;
        /* Exception from user space */
        else if (user_mode(eregs))
                regs = task_pt_regs(current);
        /* Exception from kernel and interrupts are enabled. Move to
-          kernel process stack. */
-       else if (eregs->eflags & X86_EFLAGS_IF)
-               regs = (struct pt_regs *)(eregs->rsp -= sizeof(struct pt_regs));
+          kernel process stack. */
+       else if (eregs->flags & X86_EFLAGS_IF)
+               regs = (struct pt_regs *)(eregs->sp -= sizeof(struct pt_regs));
        if (eregs != regs)
                *regs = *eregs;
        return regs;
 }
 
 /* runs on IST stack. */
-asmlinkage void __kprobes do_debug(struct pt_regs * regs,
+asmlinkage void __kprobes do_debug(struct pt_regs *regs,
                                   unsigned long error_code)
 {
-       unsigned long condition;
        struct task_struct *tsk = current;
+       unsigned long condition;
        siginfo_t info;
 
+       trace_hardirqs_fixup();
+
        get_debugreg(condition, 6);
 
+       /*
+        * The processor cleared BTF, so don't mark that we need it set.
+        */
+       clear_tsk_thread_flag(tsk, TIF_DEBUGCTLMSR);
+       tsk->thread.debugctlmsr = 0;
+
        if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code,
                                                SIGTRAP) == NOTIFY_STOP)
                return;
@@ -854,34 +920,19 @@ asmlinkage void __kprobes do_debug(struct pt_regs * regs,
 
        /* Mask out spurious debug traps due to lazy DR7 setting */
        if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
-               if (!tsk->thread.debugreg7) { 
+               if (!tsk->thread.debugreg7)
                        goto clear_dr7;
-               }
        }
 
        tsk->thread.debugreg6 = condition;
 
-       /* Mask out spurious TF errors due to lazy TF clearing */
+       /*
+        * Single-stepping through TF: make sure we ignore any events in
+        * kernel space (but re-enable TF when returning to user mode).
+        */
        if (condition & DR_STEP) {
-               /*
-                * The TF error should be masked out only if the current
-                * process is not traced and if the TRAP flag has been set
-                * previously by a tracing process (condition detected by
-                * the PT_DTRACE flag); remember that the i386 TRAP flag
-                * can be modified by the process itself in user mode,
-                * allowing programs to debug themselves without the ptrace()
-                * interface.
-                */
-                if (!user_mode(regs))
-                       goto clear_TF_reenable;
-               /*
-                * Was the TF flag set by a debugger? If so, clear it now,
-                * so that register information is correct.
-                */
-               if (tsk->ptrace & PT_DTRACE) {
-                       regs->eflags &= ~TF_MASK;
-                       tsk->ptrace &= ~PT_DTRACE;
-               }
+               if (!user_mode(regs))
+                       goto clear_TF_reenable;
        }
 
        /* Ok, finally something we can handle */
@@ -890,28 +941,26 @@ asmlinkage void __kprobes do_debug(struct pt_regs * regs,
        info.si_signo = SIGTRAP;
        info.si_errno = 0;
        info.si_code = TRAP_BRKPT;
-       info.si_addr = user_mode(regs) ? (void __user *)regs->rip : NULL;
+       info.si_addr = user_mode(regs) ? (void __user *)regs->ip : NULL;
        force_sig_info(SIGTRAP, &info, tsk);
 
 clear_dr7:
-       set_debugreg(0UL, 7);
+       set_debugreg(0, 7);
        preempt_conditional_cli(regs);
        return;
 
 clear_TF_reenable:
        set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
-       regs->eflags &= ~TF_MASK;
+       regs->flags &= ~X86_EFLAGS_TF;
        preempt_conditional_cli(regs);
+       return;
 }
 
 static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr)
 {
-       const struct exception_table_entry *fixup;
-       fixup = search_exception_tables(regs->rip);
-       if (fixup) {
-               regs->rip = fixup->fixup;
+       if (fixup_exception(regs))
                return 1;
-       }
+
        notify_die(DIE_GPF, str, regs, 0, trapnr, SIGFPE);
        /* Illegal floating point operation in the kernel */
        current->thread.trap_no = trapnr;
@@ -926,8 +975,8 @@ static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr)
  */
 asmlinkage void do_coprocessor_error(struct pt_regs *regs)
 {
-       void __user *rip = (void __user *)(regs->rip);
-       struct task_struct * task;
+       void __user *ip = (void __user *)(regs->ip);
+       struct task_struct *task;
        siginfo_t info;
        unsigned short cwd, swd;
 
@@ -946,7 +995,7 @@ asmlinkage void do_coprocessor_error(struct pt_regs *regs)
        info.si_signo = SIGFPE;
        info.si_errno = 0;
        info.si_code = __SI_FAULT;
-       info.si_addr = rip;
+       info.si_addr = ip;
        /*
         * (~cwd & swd) will mask out exceptions that are not set to unmasked
         * status.  0x3f is the exception bits in these regs, 0x200 is the
@@ -960,49 +1009,49 @@ asmlinkage void do_coprocessor_error(struct pt_regs *regs)
        cwd = get_fpu_cwd(task);
        swd = get_fpu_swd(task);
        switch (swd & ~cwd & 0x3f) {
-               case 0x000:
-               default:
-                       break;
-               case 0x001: /* Invalid Op */
-                       /*
-                        * swd & 0x240 == 0x040: Stack Underflow
-                        * swd & 0x240 == 0x240: Stack Overflow
-                        * User must clear the SF bit (0x40) if set
-                        */
-                       info.si_code = FPE_FLTINV;
-                       break;
-               case 0x002: /* Denormalize */
-               case 0x010: /* Underflow */
-                       info.si_code = FPE_FLTUND;
-                       break;
-               case 0x004: /* Zero Divide */
-                       info.si_code = FPE_FLTDIV;
-                       break;
-               case 0x008: /* Overflow */
-                       info.si_code = FPE_FLTOVF;
-                       break;
-               case 0x020: /* Precision */
-                       info.si_code = FPE_FLTRES;
-                       break;
+       case 0x000: /* No unmasked exception */
+       default: /* Multiple exceptions */
+               break;
+       case 0x001: /* Invalid Op */
+               /*
+                * swd & 0x240 == 0x040: Stack Underflow
+                * swd & 0x240 == 0x240: Stack Overflow
+                * User must clear the SF bit (0x40) if set
+                */
+               info.si_code = FPE_FLTINV;
+               break;
+       case 0x002: /* Denormalize */
+       case 0x010: /* Underflow */
+               info.si_code = FPE_FLTUND;
+               break;
+       case 0x004: /* Zero Divide */
+               info.si_code = FPE_FLTDIV;
+               break;
+       case 0x008: /* Overflow */
+               info.si_code = FPE_FLTOVF;
+               break;
+       case 0x020: /* Precision */
+               info.si_code = FPE_FLTRES;
+               break;
        }
        force_sig_info(SIGFPE, &info, task);
 }
 
 asmlinkage void bad_intr(void)
 {
-       printk("bad interrupt"); 
+       printk("bad interrupt");
 }
 
 asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
 {
-       void __user *rip = (void __user *)(regs->rip);
-       struct task_struct * task;
+       void __user *ip = (void __user *)(regs->ip);
+       struct task_struct *task;
        siginfo_t info;
        unsigned short mxcsr;
 
        conditional_sti(regs);
        if (!user_mode(regs) &&
-               kernel_math_error(regs, "kernel simd math error", 19))
+                       kernel_math_error(regs, "kernel simd math error", 19))
                return;
 
        /*
@@ -1015,7 +1064,7 @@ asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
        info.si_signo = SIGFPE;
        info.si_errno = 0;
        info.si_code = __SI_FAULT;
-       info.si_addr = rip;
+       info.si_addr = ip;
        /*
         * The SIMD FPU exceptions are handled a little differently, as there
         * is only a single status/control register.  Thus, to determine which
@@ -1024,30 +1073,30 @@ asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
         */
        mxcsr = get_fpu_mxcsr(task);
        switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
-               case 0x000:
-               default:
-                       break;
-               case 0x001: /* Invalid Op */
-                       info.si_code = FPE_FLTINV;
-                       break;
-               case 0x002: /* Denormalize */
-               case 0x010: /* Underflow */
-                       info.si_code = FPE_FLTUND;
-                       break;
-               case 0x004: /* Zero Divide */
-                       info.si_code = FPE_FLTDIV;
-                       break;
-               case 0x008: /* Overflow */
-                       info.si_code = FPE_FLTOVF;
-                       break;
-               case 0x020: /* Precision */
-                       info.si_code = FPE_FLTRES;
-                       break;
+       case 0x000:
+       default:
+               break;
+       case 0x001: /* Invalid Op */
+               info.si_code = FPE_FLTINV;
+               break;
+       case 0x002: /* Denormalize */
+       case 0x010: /* Underflow */
+               info.si_code = FPE_FLTUND;
+               break;
+       case 0x004: /* Zero Divide */
+               info.si_code = FPE_FLTDIV;
+               break;
+       case 0x008: /* Overflow */
+               info.si_code = FPE_FLTOVF;
+               break;
+       case 0x020: /* Precision */
+               info.si_code = FPE_FLTRES;
+               break;
        }
        force_sig_info(SIGFPE, &info, task);
 }
 
-asmlinkage void do_spurious_interrupt_bug(struct pt_regs * regs)
+asmlinkage void do_spurious_interrupt_bug(struct pt_regs *regs)
 {
 }
 
@@ -1060,7 +1109,7 @@ asmlinkage void __attribute__((weak)) mce_threshold_interrupt(void)
 }
 
 /*
- *  'math_state_restore()' saves the current math information in the
+ * 'math_state_restore()' saves the current math information in the
  * old math state array, and gets the new ones from the current task
  *
  * Careful.. There are problems with IBM-designed IRQ13 behaviour.
@@ -1069,66 +1118,101 @@ asmlinkage void __attribute__((weak)) mce_threshold_interrupt(void)
 asmlinkage void math_state_restore(void)
 {
        struct task_struct *me = current;
-       clts();                 /* Allow maths ops (or we recurse) */
 
-       if (!used_math())
-               init_fpu(me);
-       restore_fpu_checking(&me->thread.i387.fxsave);
+       if (!used_math()) {
+               local_irq_enable();
+               /*
+                * does a slab alloc which can sleep
+                */
+               if (init_fpu(me)) {
+                       /*
+                        * ran out of memory!
+                        */
+                       do_group_exit(SIGKILL);
+                       return;
+               }
+               local_irq_disable();
+       }
+
+       clts();                         /* Allow maths ops (or we recurse) */
+       /*
+        * Paranoid restore. send a SIGSEGV if we fail to restore the state.
+        */
+       if (unlikely(restore_fpu_checking(&me->thread.xstate->fxsave))) {
+               stts();
+               force_sig(SIGSEGV, me);
+               return;
+       }
        task_thread_info(me)->status |= TS_USEDFPU;
        me->fpu_counter++;
 }
+EXPORT_SYMBOL_GPL(math_state_restore);
 
 void __init trap_init(void)
 {
-       set_intr_gate(0,&divide_error);
-       set_intr_gate_ist(1,&debug,DEBUG_STACK);
-       set_intr_gate_ist(2,&nmi,NMI_STACK);
-       set_system_gate_ist(3,&int3,DEBUG_STACK); /* int3 can be called from all */
-       set_system_gate(4,&overflow);   /* int4 can be called from all */
-       set_intr_gate(5,&bounds);
-       set_intr_gate(6,&invalid_op);
-       set_intr_gate(7,&device_not_available);
-       set_intr_gate_ist(8,&double_fault, DOUBLEFAULT_STACK);
-       set_intr_gate(9,&coprocessor_segment_overrun);
-       set_intr_gate(10,&invalid_TSS);
-       set_intr_gate(11,&segment_not_present);
-       set_intr_gate_ist(12,&stack_segment,STACKFAULT_STACK);
-       set_intr_gate(13,&general_protection);
-       set_intr_gate(14,&page_fault);
-       set_intr_gate(15,&spurious_interrupt_bug);
-       set_intr_gate(16,&coprocessor_error);
-       set_intr_gate(17,&alignment_check);
+       set_intr_gate(0, &divide_error);
+       set_intr_gate_ist(1, &debug, DEBUG_STACK);
+       set_intr_gate_ist(2, &nmi, NMI_STACK);
+       /* int3 can be called from all */
+       set_system_gate_ist(3, &int3, DEBUG_STACK);
+       /* int4 can be called from all */
+       set_system_gate(4, &overflow);
+       set_intr_gate(5, &bounds);
+       set_intr_gate(6, &invalid_op);
+       set_intr_gate(7, &device_not_available);
+       set_intr_gate_ist(8, &double_fault, DOUBLEFAULT_STACK);
+       set_intr_gate(9, &coprocessor_segment_overrun);
+       set_intr_gate(10, &invalid_TSS);
+       set_intr_gate(11, &segment_not_present);
+       set_intr_gate_ist(12, &stack_segment, STACKFAULT_STACK);
+       set_intr_gate(13, &general_protection);
+       set_intr_gate(14, &page_fault);
+       set_intr_gate(15, &spurious_interrupt_bug);
+       set_intr_gate(16, &coprocessor_error);
+       set_intr_gate(17, &alignment_check);
 #ifdef CONFIG_X86_MCE
-       set_intr_gate_ist(18,&machine_check, MCE_STACK); 
+       set_intr_gate_ist(18, &machine_check, MCE_STACK);
 #endif
-       set_intr_gate(19,&simd_coprocessor_error);
+       set_intr_gate(19, &simd_coprocessor_error);
 
 #ifdef CONFIG_IA32_EMULATION
        set_system_gate(IA32_SYSCALL_VECTOR, ia32_syscall);
 #endif
-       
        /*
-        * Should be a barrier for any external CPU state.
+        * initialize the per thread extended state:
+        */
+       init_thread_xstate();
+       /*
+        * Should be a barrier for any external CPU state:
         */
        cpu_init();
 }
 
-
 static int __init oops_setup(char *s)
-{ 
+{
        if (!s)
                return -EINVAL;
        if (!strcmp(s, "panic"))
                panic_on_oops = 1;
        return 0;
-} 
+}
 early_param("oops", oops_setup);
 
 static int __init kstack_setup(char *s)
 {
        if (!s)
                return -EINVAL;
-       kstack_depth_to_print = simple_strtoul(s,NULL,0);
+       kstack_depth_to_print = simple_strtoul(s, NULL, 0);
        return 0;
 }
 early_param("kstack", kstack_setup);
+
+static int __init code_bytes_setup(char *s)
+{
+       code_bytes = simple_strtoul(s, NULL, 0);
+       if (code_bytes > 8192)
+               code_bytes = 8192;
+
+       return 1;
+}
+__setup("code_bytes=", code_bytes_setup);