[net-next-2.6.git] / arch / i386 / kernel / ptrace.c

/* ptrace.c */
/* By Ross Biro 1/23/92 */
/*
 * Pentium III FXSR, SSE support
 *	Gareth Hughes <gareth@valinux.com>, May 2000
 */

#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/mm.h>
#include <linux/smp.h>
#include <linux/smp_lock.h>
#include <linux/errno.h>
#include <linux/ptrace.h>
#include <linux/user.h>
#include <linux/security.h>
#include <linux/audit.h>
#include <linux/seccomp.h>
#include <linux/signal.h>

#include <asm/uaccess.h>
#include <asm/pgtable.h>
#include <asm/system.h>
#include <asm/processor.h>
#include <asm/i387.h>
#include <asm/debugreg.h>
#include <asm/ldt.h>
#include <asm/desc.h>

/*
 * does not yet catch signals sent when the child dies.
 * in exit.c or in signal.c.
 */

/* determines which flags the user has access to. */
/* 1 = access 0 = no access */
#define FLAG_MASK 0x00044dd5

/* set's the trap flag. */
#define TRAP_FLAG 0x100

/*
 * Offset of eflags on child stack..
 */
#define EFL_OFFSET ((EFL-2)*4-sizeof(struct pt_regs))

static inline struct pt_regs *get_child_regs(struct task_struct *task)
{
	void *stack_top = (void *)task->thread.esp0;
	return stack_top - sizeof(struct pt_regs);
}

/*
 * this routine will get a word off of the processes privileged stack. 
 * the offset is how far from the base addr as stored in the TSS.  
 * this routine assumes that all the privileged stacks are in our
 * data space.
 */   
static inline int get_stack_long(struct task_struct *task, int offset)
{
	unsigned char *stack;

	stack = (unsigned char *)task->thread.esp0;
	stack += offset;
	return (*((int *)stack));
}

/*
 * this routine will put a word on the processes privileged stack. 
 * the offset is how far from the base addr as stored in the TSS.  
 * this routine assumes that all the privileged stacks are in our
 * data space.
 */
static inline int put_stack_long(struct task_struct *task, int offset,
	unsigned long data)
{
	unsigned char * stack;

	stack = (unsigned char *) task->thread.esp0;
	stack += offset;
	*(unsigned long *) stack = data;
	return 0;
}

static int putreg(struct task_struct *child,
	unsigned long regno, unsigned long value)
{
	switch (regno >> 2) {
		case FS:
			if (value && (value & 3) != 3)
				return -EIO;
			child->thread.fs = value;
			return 0;
		case GS:
			if (value && (value & 3) != 3)
				return -EIO;
			child->thread.gs = value;
			return 0;
		case DS:
		case ES:
			if (value && (value & 3) != 3)
				return -EIO;
			value &= 0xffff;
			break;
		case SS:
		case CS:
			if ((value & 3) != 3)
				return -EIO;
			value &= 0xffff;
			break;
		case EFL:
			value &= FLAG_MASK;
			value |= get_stack_long(child, EFL_OFFSET) & ~FLAG_MASK;
			break;
	}
	if (regno > GS*4)
		regno -= 2*4;
	put_stack_long(child, regno - sizeof(struct pt_regs), value);
	return 0;
}

static unsigned long getreg(struct task_struct *child,
	unsigned long regno)
{
	unsigned long retval = ~0UL;

	switch (regno >> 2) {
		case FS:
			retval = child->thread.fs;
			break;
		case GS:
			retval = child->thread.gs;
			break;
		case DS:
		case ES:
		case SS:
		case CS:
			retval = 0xffff;
			/* fall through */
		default:
			if (regno > GS*4)
				regno -= 2*4;
			regno = regno - sizeof(struct pt_regs);
			retval &= get_stack_long(child, regno);
	}
	return retval;
}

#define LDT_SEGMENT 4

static unsigned long convert_eip_to_linear(struct task_struct *child, struct pt_regs *regs)
{
	unsigned long addr, seg;

	addr = regs->eip;
	seg = regs->xcs & 0xffff;
	if (regs->eflags & VM_MASK) {
		addr = (addr & 0xffff) + (seg << 4);
		return addr;
	}

	/*
	 * We'll assume that the code segments in the GDT
	 * are all zero-based. That is largely true: the
	 * TLS segments are used for data, and the PNPBIOS
	 * and APM bios ones we just ignore here.
	 */
	if (seg & LDT_SEGMENT) {
		u32 *desc;
		unsigned long base;

		down(&child->mm->context.sem);
		desc = child->mm->context.ldt + (seg & ~7);
		base = (desc[0] >> 16) | ((desc[1] & 0xff) << 16) | (desc[1] & 0xff000000);

		/* 16-bit code segment? */
		if (!((desc[1] >> 22) & 1))
			addr &= 0xffff;
		addr += base;
		up(&child->mm->context.sem);
	}
	return addr;
}

static inline int is_at_popf(struct task_struct *child, struct pt_regs *regs)
{
	int i, copied;
	unsigned char opcode[16];
	unsigned long addr = convert_eip_to_linear(child, regs);

	copied = access_process_vm(child, addr, opcode, sizeof(opcode), 0);
	for (i = 0; i < copied; i++) {
		switch (opcode[i]) {
		/* popf */
		case 0x9d:
			return 1;
		/* opcode and address size prefixes */
		case 0x66: case 0x67:
			continue;
		/* irrelevant prefixes (segment overrides and repeats) */
		case 0x26: case 0x2e:
		case 0x36: case 0x3e:
		case 0x64: case 0x65:
		case 0xf0: case 0xf2: case 0xf3:
			continue;

		/*
		 * pushf: NOTE! We should probably not let
		 * the user see the TF bit being set. But
		 * it's more pain than it's worth to avoid
		 * it, and a debugger could emulate this
		 * all in user space if it _really_ cares.
		 */
		case 0x9c:
		default:
			return 0;
		}
	}
	return 0;
}

static void set_singlestep(struct task_struct *child)
{
	struct pt_regs *regs = get_child_regs(child);

	/*
	 * Always set TIF_SINGLESTEP - this guarantees that 
	 * we single-step system calls etc..  This will also
	 * cause us to set TF when returning to user mode.
	 */
	set_tsk_thread_flag(child, TIF_SINGLESTEP);

	/*
	 * If TF was already set, don't do anything else
	 */
	if (regs->eflags & TRAP_FLAG)
		return;

	/* Set TF on the kernel stack.. */
	regs->eflags |= TRAP_FLAG;

	/*
	 * ..but if TF is changed by the instruction we will trace,
	 * don't mark it as being "us" that set it, so that we
	 * won't clear it by hand later.
	 */
	if (is_at_popf(child, regs))
		return;
	
	child->ptrace |= PT_DTRACE;
}

static void clear_singlestep(struct task_struct *child)
{
	/* Always clear TIF_SINGLESTEP... */
	clear_tsk_thread_flag(child, TIF_SINGLESTEP);

	/* But touch TF only if it was set by us.. */
	if (child->ptrace & PT_DTRACE) {
		struct pt_regs *regs = get_child_regs(child);
		regs->eflags &= ~TRAP_FLAG;
		child->ptrace &= ~PT_DTRACE;
	}
}

/*
 * Called by kernel/ptrace.c when detaching..
 *
 * Make sure the single step bit is not set.
 */
void ptrace_disable(struct task_struct *child)
{ 
	clear_singlestep(child);
	clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
	clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
}

/*
 * Perform get_thread_area on behalf of the traced child.
 */
static int
ptrace_get_thread_area(struct task_struct *child,
		       int idx, struct user_desc __user *user_desc)
{
	struct user_desc info;
	struct desc_struct *desc;

/*
 * Get the current Thread-Local Storage area:
 */

#define GET_BASE(desc) ( \
	(((desc)->a >> 16) & 0x0000ffff) | \
	(((desc)->b << 16) & 0x00ff0000) | \
	( (desc)->b        & 0xff000000)   )

#define GET_LIMIT(desc) ( \
	((desc)->a & 0x0ffff) | \
	 ((desc)->b & 0xf0000) )

#define GET_32BIT(desc)		(((desc)->b >> 22) & 1)
#define GET_CONTENTS(desc)	(((desc)->b >> 10) & 3)
#define GET_WRITABLE(desc)	(((desc)->b >>  9) & 1)
#define GET_LIMIT_PAGES(desc)	(((desc)->b >> 23) & 1)
#define GET_PRESENT(desc)	(((desc)->b >> 15) & 1)
#define GET_USEABLE(desc)	(((desc)->b >> 20) & 1)

	if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
		return -EINVAL;

	desc = child->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;

	info.entry_number = idx;
	info.base_addr = GET_BASE(desc);
	info.limit = GET_LIMIT(desc);
	info.seg_32bit = GET_32BIT(desc);
	info.contents = GET_CONTENTS(desc);
	info.read_exec_only = !GET_WRITABLE(desc);
	info.limit_in_pages = GET_LIMIT_PAGES(desc);
	info.seg_not_present = !GET_PRESENT(desc);
	info.useable = GET_USEABLE(desc);

	if (copy_to_user(user_desc, &info, sizeof(info)))
		return -EFAULT;

	return 0;
}

/*
 * Perform set_thread_area on behalf of the traced child.
 */
static int
ptrace_set_thread_area(struct task_struct *child,
		       int idx, struct user_desc __user *user_desc)
{
	struct user_desc info;
	struct desc_struct *desc;

	if (copy_from_user(&info, user_desc, sizeof(info)))
		return -EFAULT;

	if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX)
		return -EINVAL;

	desc = child->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
	if (LDT_empty(&info)) {
		desc->a = 0;
		desc->b = 0;
	} else {
		desc->a = LDT_entry_a(&info);
		desc->b = LDT_entry_b(&info);
	}

	return 0;
}

long arch_ptrace(struct task_struct *child, long request, long addr, long data)
{
	struct user * dummy = NULL;
	int i, ret;
	unsigned long __user *datap = (unsigned long __user *)data;

	switch (request) {
	/* when I and D space are separate, these will need to be fixed. */
	case PTRACE_PEEKTEXT: /* read word at location addr. */ 
	case PTRACE_PEEKDATA: {
		unsigned long tmp;
		int copied;

		copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 0);
		ret = -EIO;
		if (copied != sizeof(tmp))
			break;
		ret = put_user(tmp, datap);
		break;
	}

	/* read the word at location addr in the USER area. */
	case PTRACE_PEEKUSR: {
		unsigned long tmp;

		ret = -EIO;
		if ((addr & 3) || addr < 0 || 
		    addr > sizeof(struct user) - 3)
			break;

		tmp = 0;  /* Default return condition */
		if(addr < FRAME_SIZE*sizeof(long))
			tmp = getreg(child, addr);
		if(addr >= (long) &dummy->u_debugreg[0] &&
		   addr <= (long) &dummy->u_debugreg[7]){
			addr -= (long) &dummy->u_debugreg[0];
			addr = addr >> 2;
			tmp = child->thread.debugreg[addr];
		}
		ret = put_user(tmp, datap);
		break;
	}

	/* when I and D space are separate, this will have to be fixed. */
	case PTRACE_POKETEXT: /* write the word at location addr. */
	case PTRACE_POKEDATA:
		ret = 0;
		if (access_process_vm(child, addr, &data, sizeof(data), 1) == sizeof(data))
			break;
		ret = -EIO;
		break;

	case PTRACE_POKEUSR: /* write the word at location addr in the USER area */
		ret = -EIO;
		if ((addr & 3) || addr < 0 || 
		    addr > sizeof(struct user) - 3)
			break;

		if (addr < FRAME_SIZE*sizeof(long)) {
			ret = putreg(child, addr, data);
			break;
		}
		/* We need to be very careful here.  We implicitly
		   want to modify a portion of the task_struct, and we
		   have to be selective about what portions we allow someone
		   to modify. */

		  ret = -EIO;
		  if(addr >= (long) &dummy->u_debugreg[0] &&
		     addr <= (long) &dummy->u_debugreg[7]){

			  if(addr == (long) &dummy->u_debugreg[4]) break;
			  if(addr == (long) &dummy->u_debugreg[5]) break;
			  if(addr < (long) &dummy->u_debugreg[4] &&
			     ((unsigned long) data) >= TASK_SIZE-3) break;
			  
			  /* Sanity-check data. Take one half-byte at once with
			   * check = (val >> (16 + 4*i)) & 0xf. It contains the
			   * R/Wi and LENi bits; bits 0 and 1 are R/Wi, and bits
			   * 2 and 3 are LENi. Given a list of invalid values,
			   * we do mask |= 1 << invalid_value, so that
			   * (mask >> check) & 1 is a correct test for invalid
			   * values.
			   *
			   * R/Wi contains the type of the breakpoint /
			   * watchpoint, LENi contains the length of the watched
			   * data in the watchpoint case.
			   *
			   * The invalid values are:
			   * - LENi == 0x10 (undefined), so mask |= 0x0f00.
			   * - R/Wi == 0x10 (break on I/O reads or writes), so
			   *   mask |= 0x4444.
			   * - R/Wi == 0x00 && LENi != 0x00, so we have mask |=
			   *   0x1110.
			   *
			   * Finally, mask = 0x0f00 | 0x4444 | 0x1110 == 0x5f54.
			   *
			   * See the Intel Manual "System Programming Guide",
			   * 15.2.4
			   *
			   * Note that LENi == 0x10 is defined on x86_64 in long
			   * mode (i.e. even for 32-bit userspace software, but
			   * 64-bit kernel), so the x86_64 mask value is 0x5454.
			   * See the AMD manual no. 24593 (AMD64 System
			   * Programming)*/

			  if(addr == (long) &dummy->u_debugreg[7]) {
				  data &= ~DR_CONTROL_RESERVED;
				  for(i=0; i<4; i++)
					  if ((0x5f54 >> ((data >> (16 + 4*i)) & 0xf)) & 1)
						  goto out_tsk;
			  }

			  addr -= (long) &dummy->u_debugreg;
			  addr = addr >> 2;
			  child->thread.debugreg[addr] = data;
			  ret = 0;
		  }
		  break;

	case PTRACE_SYSEMU: /* continue and stop at next syscall, which will not be executed */
	case PTRACE_SYSCALL:	/* continue and stop at next (return from) syscall */
	case PTRACE_CONT:	/* restart after signal. */
		ret = -EIO;
		if (!valid_signal(data))
			break;
		if (request == PTRACE_SYSEMU) {
			set_tsk_thread_flag(child, TIF_SYSCALL_EMU);
			clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
		} else if (request == PTRACE_SYSCALL) {
			set_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
			clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
		} else {
			clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
			clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
		}
		child->exit_code = data;
		/* make sure the single step bit is not set. */
		clear_singlestep(child);
		wake_up_process(child);
		ret = 0;
		break;

/*
 * make the child exit.  Best I can do is send it a sigkill. 
 * perhaps it should be put in the status that it wants to 
 * exit.
 */
	case PTRACE_KILL:
		ret = 0;
		if (child->exit_state == EXIT_ZOMBIE)	/* already dead */
			break;
		child->exit_code = SIGKILL;
		/* make sure the single step bit is not set. */
		clear_singlestep(child);
		wake_up_process(child);
		break;

	case PTRACE_SYSEMU_SINGLESTEP: /* Same as SYSEMU, but singlestep if not syscall */
	case PTRACE_SINGLESTEP:	/* set the trap flag. */
		ret = -EIO;
		if (!valid_signal(data))
			break;

		if (request == PTRACE_SYSEMU_SINGLESTEP)
			set_tsk_thread_flag(child, TIF_SYSCALL_EMU);
		else
			clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);

		clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
		set_singlestep(child);
		child->exit_code = data;
		/* give it a chance to run. */
		wake_up_process(child);
		ret = 0;
		break;

	case PTRACE_DETACH:
		/* detach a process that was attached. */
		ret = ptrace_detach(child, data);
		break;

	case PTRACE_GETREGS: { /* Get all gp regs from the child. */
	  	if (!access_ok(VERIFY_WRITE, datap, FRAME_SIZE*sizeof(long))) {
			ret = -EIO;
			break;
		}
		for ( i = 0; i < FRAME_SIZE*sizeof(long); i += sizeof(long) ) {
			__put_user(getreg(child, i), datap);
			datap++;
		}
		ret = 0;
		break;
	}

	case PTRACE_SETREGS: { /* Set all gp regs in the child. */
		unsigned long tmp;
	  	if (!access_ok(VERIFY_READ, datap, FRAME_SIZE*sizeof(long))) {
			ret = -EIO;
			break;
		}
		for ( i = 0; i < FRAME_SIZE*sizeof(long); i += sizeof(long) ) {
			__get_user(tmp, datap);
			putreg(child, i, tmp);
			datap++;
		}
		ret = 0;
		break;
	}

	case PTRACE_GETFPREGS: { /* Get the child FPU state. */
		if (!access_ok(VERIFY_WRITE, datap,
			       sizeof(struct user_i387_struct))) {
			ret = -EIO;
			break;
		}
		ret = 0;
		if (!tsk_used_math(child))
			init_fpu(child);
		get_fpregs((struct user_i387_struct __user *)data, child);
		break;
	}

	case PTRACE_SETFPREGS: { /* Set the child FPU state. */
		if (!access_ok(VERIFY_READ, datap,
			       sizeof(struct user_i387_struct))) {
			ret = -EIO;
			break;
		}
		set_stopped_child_used_math(child);
		set_fpregs(child, (struct user_i387_struct __user *)data);
		ret = 0;
		break;
	}

	case PTRACE_GETFPXREGS: { /* Get the child extended FPU state. */
		if (!access_ok(VERIFY_WRITE, datap,
			       sizeof(struct user_fxsr_struct))) {
			ret = -EIO;
			break;
		}
		if (!tsk_used_math(child))
			init_fpu(child);
		ret = get_fpxregs((struct user_fxsr_struct __user *)data, child);
		break;
	}

	case PTRACE_SETFPXREGS: { /* Set the child extended FPU state. */
		if (!access_ok(VERIFY_READ, datap,
			       sizeof(struct user_fxsr_struct))) {
			ret = -EIO;
			break;
		}
		set_stopped_child_used_math(child);
		ret = set_fpxregs(child, (struct user_fxsr_struct __user *)data);
		break;
	}

	case PTRACE_GET_THREAD_AREA:
		ret = ptrace_get_thread_area(child, addr,
					(struct user_desc __user *) data);
		break;

	case PTRACE_SET_THREAD_AREA:
		ret = ptrace_set_thread_area(child, addr,
					(struct user_desc __user *) data);
		break;

	default:
		ret = ptrace_request(child, request, addr, data);
		break;
	}
 out_tsk:
	return ret;
}

void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code)
{
	struct siginfo info;

	tsk->thread.trap_no = 1;
	tsk->thread.error_code = error_code;

	memset(&info, 0, sizeof(info));
	info.si_signo = SIGTRAP;
	info.si_code = TRAP_BRKPT;

	/* User-mode eip? */
	info.si_addr = user_mode_vm(regs) ? (void __user *) regs->eip : NULL;

	/* Send us the fakey SIGTRAP */
	force_sig_info(SIGTRAP, &info, tsk);
}

/* notification of system call entry/exit
 * - triggered by current->work.syscall_trace
 */
__attribute__((regparm(3)))
int do_syscall_trace(struct pt_regs *regs, int entryexit)
{
	int is_sysemu = test_thread_flag(TIF_SYSCALL_EMU);
	/*
	 * With TIF_SYSCALL_EMU set we want to ignore TIF_SINGLESTEP for syscall
	 * interception
	 */
	int is_singlestep = !is_sysemu && test_thread_flag(TIF_SINGLESTEP);
	int ret = 0;

	/* do the secure computing check first */
	if (!entryexit)
		secure_computing(regs->orig_eax);

	if (unlikely(current->audit_context)) {
		if (entryexit)
			audit_syscall_exit(current, AUDITSC_RESULT(regs->eax),
						regs->eax);
		/* Debug traps, when using PTRACE_SINGLESTEP, must be sent only
		 * on the syscall exit path. Normally, when TIF_SYSCALL_AUDIT is
		 * not used, entry.S will call us only on syscall exit, not
		 * entry; so when TIF_SYSCALL_AUDIT is used we must avoid
		 * calling send_sigtrap() on syscall entry.
		 *
		 * Note that when PTRACE_SYSEMU_SINGLESTEP is used,
		 * is_singlestep is false, despite his name, so we will still do
		 * the correct thing.
		 */
		else if (is_singlestep)
			goto out;
	}

	if (!(current->ptrace & PT_PTRACED))
		goto out;

	/* If a process stops on the 1st tracepoint with SYSCALL_TRACE
	 * and then is resumed with SYSEMU_SINGLESTEP, it will come in
	 * here. We have to check this and return */
	if (is_sysemu && entryexit)
		return 0;

	/* Fake a debug trap */
	if (is_singlestep)
		send_sigtrap(current, regs, 0);

 	if (!test_thread_flag(TIF_SYSCALL_TRACE) && !is_sysemu)
		goto out;

	/* the 0x80 provides a way for the tracing parent to distinguish
	   between a syscall stop and SIGTRAP delivery */
	/* Note that the debugger could change the result of test_thread_flag!*/
	ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD) ? 0x80:0));

	/*
	 * this isn't the same as continuing with a signal, but it will do
	 * for normal use.  strace only continues with a signal if the
	 * stopping signal is not SIGTRAP.  -brl
	 */
	if (current->exit_code) {
		send_sig(current->exit_code, current, 1);
		current->exit_code = 0;
	}
	ret = is_sysemu;
out:
	if (unlikely(current->audit_context) && !entryexit)
		audit_syscall_entry(current, AUDIT_ARCH_I386, regs->orig_eax,
				    regs->ebx, regs->ecx, regs->edx, regs->esi);
	if (ret == 0)
		return 0;

	regs->orig_eax = -1; /* force skip of syscall restarting */
	if (unlikely(current->audit_context))
		audit_syscall_exit(current, AUDITSC_RESULT(regs->eax),
				regs->eax);
	return 1;
}
Commit	Line	Data
1da177e4 LT	1	/* ptrace.c */
	2	/* By Ross Biro 1/23/92 */
	3	/*
	4	* Pentium III FXSR, SSE support
	5	* Gareth Hughes <gareth@valinux.com>, May 2000
	6	*/
	7
	8	#include <linux/kernel.h>
	9	#include <linux/sched.h>
	10	#include <linux/mm.h>
	11	#include <linux/smp.h>
	12	#include <linux/smp_lock.h>
	13	#include <linux/errno.h>
	14	#include <linux/ptrace.h>
	15	#include <linux/user.h>
	16	#include <linux/security.h>
	17	#include <linux/audit.h>
	18	#include <linux/seccomp.h>
7ed20e1a	19	#include <linux/signal.h>
1da177e4 LT	20
	21	#include <asm/uaccess.h>
	22	#include <asm/pgtable.h>
	23	#include <asm/system.h>
	24	#include <asm/processor.h>
	25	#include <asm/i387.h>
	26	#include <asm/debugreg.h>
	27	#include <asm/ldt.h>
	28	#include <asm/desc.h>
	29
	30	/*
	31	* does not yet catch signals sent when the child dies.
	32	* in exit.c or in signal.c.
	33	*/
	34
	35	/* determines which flags the user has access to. */
	36	/* 1 = access 0 = no access */
	37	#define FLAG_MASK 0x00044dd5
	38
	39	/* set's the trap flag. */
	40	#define TRAP_FLAG 0x100
	41
	42	/*
	43	* Offset of eflags on child stack..
	44	*/
	45	#define EFL_OFFSET ((EFL-2)*4-sizeof(struct pt_regs))
	46
	47	static inline struct pt_regs get_child_regs(struct task_struct task)
	48	{
	49	void stack_top = (void )task->thread.esp0;
	50	return stack_top - sizeof(struct pt_regs);
	51	}
	52
	53	/*
	54	* this routine will get a word off of the processes privileged stack.
	55	* the offset is how far from the base addr as stored in the TSS.
	56	* this routine assumes that all the privileged stacks are in our
	57	* data space.
	58	*/
	59	static inline int get_stack_long(struct task_struct *task, int offset)
	60	{
	61	unsigned char *stack;
	62
	63	stack = (unsigned char *)task->thread.esp0;
	64	stack += offset;
	65	return (((int )stack));
	66	}
	67
	68	/*
	69	* this routine will put a word on the processes privileged stack.
	70	* the offset is how far from the base addr as stored in the TSS.
	71	* this routine assumes that all the privileged stacks are in our
	72	* data space.
	73	*/
	74	static inline int put_stack_long(struct task_struct *task, int offset,
	75	unsigned long data)
	76	{
	77	unsigned char * stack;
	78
	79	stack = (unsigned char *) task->thread.esp0;
	80	stack += offset;
	81	(unsigned long ) stack = data;
	82	return 0;
	83	}
84
85	static int putreg(struct task_struct *child,
86	unsigned long regno, unsigned long value)
87	{
88	switch (regno >> 2) {
89	case FS:
90	if (value && (value & 3) != 3)
91	return -EIO;
92	child->thread.fs = value;
93	return 0;
94	case GS:
95	if (value && (value & 3) != 3)
96	return -EIO;
97	child->thread.gs = value;
98	return 0;
99	case DS:
100	case ES:
101	if (value && (value & 3) != 3)
102	return -EIO;
103	value &= 0xffff;
104	break;
105	case SS:
106	case CS:
107	if ((value & 3) != 3)
108	return -EIO;
109	value &= 0xffff;
110	break;
111	case EFL:
112	value &= FLAG_MASK;
113	value \|= get_stack_long(child, EFL_OFFSET) & ~FLAG_MASK;
114	break;
115	}
116	if (regno > GS*4)
117	regno -= 2*4;
118	put_stack_long(child, regno - sizeof(struct pt_regs), value);
119	return 0;
120	}
121
122	static unsigned long getreg(struct task_struct *child,
123	unsigned long regno)
124	{
125	unsigned long retval = ~0UL;
126
127	switch (regno >> 2) {
128	case FS:
129	retval = child->thread.fs;
130	break;
131	case GS:
132	retval = child->thread.gs;
133	break;
134	case DS:
135	case ES:
136	case SS:
137	case CS:
138	retval = 0xffff;
139	/* fall through */
140	default:
141	if (regno > GS*4)
142	regno -= 2*4;
143	regno = regno - sizeof(struct pt_regs);
144	retval &= get_stack_long(child, regno);
145	}
146	return retval;
147	}
148
149	#define LDT_SEGMENT 4
150
151	static unsigned long convert_eip_to_linear(struct task_struct child, struct pt_regs regs)
152	{
153	unsigned long addr, seg;
154
155	addr = regs->eip;
156	seg = regs->xcs & 0xffff;
157	if (regs->eflags & VM_MASK) {
158	addr = (addr & 0xffff) + (seg << 4);
159	return addr;
160	}
161
162	/*
163	* We'll assume that the code segments in the GDT
164	* are all zero-based. That is largely true: the
165	* TLS segments are used for data, and the PNPBIOS
166	* and APM bios ones we just ignore here.
167	*/
168	if (seg & LDT_SEGMENT) {
169	u32 *desc;
170	unsigned long base;
171
172	down(&child->mm->context.sem);
173	desc = child->mm->context.ldt + (seg & ~7);
174	base = (desc[0] >> 16) \| ((desc[1] & 0xff) << 16) \| (desc[1] & 0xff000000);
175
176	/* 16-bit code segment? */
177	if (!((desc[1] >> 22) & 1))
178	addr &= 0xffff;
179	addr += base;
180	up(&child->mm->context.sem);
181	}
182	return addr;
183	}
184
185	static inline int is_at_popf(struct task_struct child, struct pt_regs regs)
186	{
187	int i, copied;
188	unsigned char opcode[16];
189	unsigned long addr = convert_eip_to_linear(child, regs);
190
191	copied = access_process_vm(child, addr, opcode, sizeof(opcode), 0);
192	for (i = 0; i < copied; i++) {
193	switch (opcode[i]) {
194	/* popf */
195	case 0x9d:
196	return 1;
197	/* opcode and address size prefixes */
198	case 0x66: case 0x67:
199	continue;
200	/* irrelevant prefixes (segment overrides and repeats) */
201	case 0x26: case 0x2e:
202	case 0x36: case 0x3e:
203	case 0x64: case 0x65:
204	case 0xf0: case 0xf2: case 0xf3:
205	continue;
206
207	/*
208	* pushf: NOTE! We should probably not let
209	* the user see the TF bit being set. But
210	* it's more pain than it's worth to avoid
211	* it, and a debugger could emulate this
212	* all in user space if it _really_ cares.
213	*/
214	case 0x9c:
215	default:
216	return 0;
217	}
218	}
219	return 0;
220	}
221
222	static void set_singlestep(struct task_struct *child)
223	{
224	struct pt_regs *regs = get_child_regs(child);
225
226	/*
227	* Always set TIF_SINGLESTEP - this guarantees that
228	* we single-step system calls etc.. This will also
229	* cause us to set TF when returning to user mode.
230	*/
231	set_tsk_thread_flag(child, TIF_SINGLESTEP);
232
233	/*
234	* If TF was already set, don't do anything else
235	*/
236	if (regs->eflags & TRAP_FLAG)
237	return;
238
239	/* Set TF on the kernel stack.. */
240	regs->eflags \|= TRAP_FLAG;
241
242	/*
243	* ..but if TF is changed by the instruction we will trace,
244	* don't mark it as being "us" that set it, so that we
245	* won't clear it by hand later.
246	*/
247	if (is_at_popf(child, regs))
248	return;
249
250	child->ptrace \|= PT_DTRACE;
251	}
252
253	static void clear_singlestep(struct task_struct *child)
254	{
255	/* Always clear TIF_SINGLESTEP... */
256	clear_tsk_thread_flag(child, TIF_SINGLESTEP);
257
258	/* But touch TF only if it was set by us.. */
259	if (child->ptrace & PT_DTRACE) {
260	struct pt_regs *regs = get_child_regs(child);
261	regs->eflags &= ~TRAP_FLAG;
262	child->ptrace &= ~PT_DTRACE;
263	}
264	}
265
266	/*
267	* Called by kernel/ptrace.c when detaching..
268	*
269	* Make sure the single step bit is not set.
270	*/
271	void ptrace_disable(struct task_struct *child)
272	{
273	clear_singlestep(child);
ab1c23c2 BS	274	clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
ab1c23c2 BS	275	clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
1da177e4 LT	276	}
	277
	278	/*
	279	* Perform get_thread_area on behalf of the traced child.
	280	*/
	281	static int
	282	ptrace_get_thread_area(struct task_struct *child,
	283	int idx, struct user_desc __user *user_desc)
	284	{
	285	struct user_desc info;
	286	struct desc_struct *desc;
	287
	288	/*
	289	* Get the current Thread-Local Storage area:
	290	*/
	291
	292	#define GET_BASE(desc) ( \
	293	(((desc)->a >> 16) & 0x0000ffff) \| \
	294	(((desc)->b << 16) & 0x00ff0000) \| \
	295	( (desc)->b & 0xff000000) )
	296
	297	#define GET_LIMIT(desc) ( \
	298	((desc)->a & 0x0ffff) \| \
	299	((desc)->b & 0xf0000) )
	300
	301	#define GET_32BIT(desc) (((desc)->b >> 22) & 1)
	302	#define GET_CONTENTS(desc) (((desc)->b >> 10) & 3)
	303	#define GET_WRITABLE(desc) (((desc)->b >> 9) & 1)
	304	#define GET_LIMIT_PAGES(desc) (((desc)->b >> 23) & 1)
	305	#define GET_PRESENT(desc) (((desc)->b >> 15) & 1)
	306	#define GET_USEABLE(desc) (((desc)->b >> 20) & 1)
	307
	308	if (idx < GDT_ENTRY_TLS_MIN \|\| idx > GDT_ENTRY_TLS_MAX)
	309	return -EINVAL;
	310
	311	desc = child->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
	312
	313	info.entry_number = idx;
	314	info.base_addr = GET_BASE(desc);
	315	info.limit = GET_LIMIT(desc);
	316	info.seg_32bit = GET_32BIT(desc);
	317	info.contents = GET_CONTENTS(desc);
	318	info.read_exec_only = !GET_WRITABLE(desc);
	319	info.limit_in_pages = GET_LIMIT_PAGES(desc);
	320	info.seg_not_present = !GET_PRESENT(desc);
	321	info.useable = GET_USEABLE(desc);
	322
	323	if (copy_to_user(user_desc, &info, sizeof(info)))
	324	return -EFAULT;
	325
	326	return 0;
	327	}
	328
	329	/*
	330	* Perform set_thread_area on behalf of the traced child.
	331	*/
	332	static int
	333	ptrace_set_thread_area(struct task_struct *child,
	334	int idx, struct user_desc __user *user_desc)
	335	{
	336	struct user_desc info;
	337	struct desc_struct *desc;
	338
	339	if (copy_from_user(&info, user_desc, sizeof(info)))
340	return -EFAULT;
341
342	if (idx < GDT_ENTRY_TLS_MIN \|\| idx > GDT_ENTRY_TLS_MAX)
343	return -EINVAL;
344
345	desc = child->thread.tls_array + idx - GDT_ENTRY_TLS_MIN;
346	if (LDT_empty(&info)) {
347	desc->a = 0;
348	desc->b = 0;
349	} else {
350	desc->a = LDT_entry_a(&info);
351	desc->b = LDT_entry_b(&info);
352	}
353
354	return 0;
355	}
356
481bed45	357	long arch_ptrace(struct task_struct *child, long request, long addr, long data)
1da177e4	358	{
1da177e4 LT	359	struct user * dummy = NULL;
	360	int i, ret;
	361	unsigned long __user datap = (unsigned long __user )data;
	362
1da177e4 LT	363	switch (request) {
	364	/* when I and D space are separate, these will need to be fixed. */
	365	case PTRACE_PEEKTEXT: /* read word at location addr. */
	366	case PTRACE_PEEKDATA: {
	367	unsigned long tmp;
	368	int copied;
	369
	370	copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 0);
	371	ret = -EIO;
	372	if (copied != sizeof(tmp))
	373	break;
	374	ret = put_user(tmp, datap);
	375	break;
	376	}
	377
	378	/* read the word at location addr in the USER area. */
	379	case PTRACE_PEEKUSR: {
	380	unsigned long tmp;
	381
	382	ret = -EIO;
	383	if ((addr & 3) \|\| addr < 0 \|\|
	384	addr > sizeof(struct user) - 3)
	385	break;
	386
	387	tmp = 0; /* Default return condition */
	388	if(addr < FRAME_SIZE*sizeof(long))
	389	tmp = getreg(child, addr);
	390	if(addr >= (long) &dummy->u_debugreg[0] &&
	391	addr <= (long) &dummy->u_debugreg[7]){
	392	addr -= (long) &dummy->u_debugreg[0];
	393	addr = addr >> 2;
	394	tmp = child->thread.debugreg[addr];
	395	}
	396	ret = put_user(tmp, datap);
	397	break;
	398	}
	399
	400	/* when I and D space are separate, this will have to be fixed. */
	401	case PTRACE_POKETEXT: /* write the word at location addr. */
	402	case PTRACE_POKEDATA:
	403	ret = 0;
	404	if (access_process_vm(child, addr, &data, sizeof(data), 1) == sizeof(data))
	405	break;
	406	ret = -EIO;
	407	break;
	408
	409	case PTRACE_POKEUSR: /* write the word at location addr in the USER area */
	410	ret = -EIO;
	411	if ((addr & 3) \|\| addr < 0 \|\|
	412	addr > sizeof(struct user) - 3)
	413	break;
	414
	415	if (addr < FRAME_SIZE*sizeof(long)) {
	416	ret = putreg(child, addr, data);
	417	break;
	418	}
	419	/* We need to be very careful here. We implicitly
	420	want to modify a portion of the task_struct, and we
	421	have to be selective about what portions we allow someone
	422	to modify. */
	423
	424	ret = -EIO;
	425	if(addr >= (long) &dummy->u_debugreg[0] &&
	426	addr <= (long) &dummy->u_debugreg[7]){
427
428	if(addr == (long) &dummy->u_debugreg[4]) break;
429	if(addr == (long) &dummy->u_debugreg[5]) break;
430	if(addr < (long) &dummy->u_debugreg[4] &&
431	((unsigned long) data) >= TASK_SIZE-3) break;
432
433	/* Sanity-check data. Take one half-byte at once with
434	* check = (val >> (16 + 4*i)) & 0xf. It contains the
435	* R/Wi and LENi bits; bits 0 and 1 are R/Wi, and bits
436	* 2 and 3 are LENi. Given a list of invalid values,
437	* we do mask \|= 1 << invalid_value, so that
438	* (mask >> check) & 1 is a correct test for invalid
439	* values.
440	*
441	* R/Wi contains the type of the breakpoint /
442	* watchpoint, LENi contains the length of the watched
443	* data in the watchpoint case.
444	*
445	* The invalid values are:
446	* - LENi == 0x10 (undefined), so mask \|= 0x0f00.
447	* - R/Wi == 0x10 (break on I/O reads or writes), so
448	* mask \|= 0x4444.
449	* - R/Wi == 0x00 && LENi != 0x00, so we have mask \|=
450	* 0x1110.
451	*
452	* Finally, mask = 0x0f00 \| 0x4444 \| 0x1110 == 0x5f54.
453	*
454	* See the Intel Manual "System Programming Guide",
455	* 15.2.4
456	*
457	* Note that LENi == 0x10 is defined on x86_64 in long
458	* mode (i.e. even for 32-bit userspace software, but
459	* 64-bit kernel), so the x86_64 mask value is 0x5454.
460	* See the AMD manual no. 24593 (AMD64 System
461	* Programming)*/
462
463	if(addr == (long) &dummy->u_debugreg[7]) {
464	data &= ~DR_CONTROL_RESERVED;
465	for(i=0; i<4; i++)
466	if ((0x5f54 >> ((data >> (16 + 4*i)) & 0xf)) & 1)
467	goto out_tsk;
468	}
469
470	addr -= (long) &dummy->u_debugreg;
471	addr = addr >> 2;
472	child->thread.debugreg[addr] = data;
473	ret = 0;
474	}
475	break;
476
ed75e8d5	477	case PTRACE_SYSEMU: /* continue and stop at next syscall, which will not be executed */
1da177e4 LT	478	case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */
	479	case PTRACE_CONT: /* restart after signal. */
	480	ret = -EIO;
7ed20e1a	481	if (!valid_signal(data))
1da177e4	482	break;
ed75e8d5 LV	483	if (request == PTRACE_SYSEMU) {
ed75e8d5 LV	484	set_tsk_thread_flag(child, TIF_SYSCALL_EMU);
c8c86cec BS	485	clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
c8c86cec BS	486	} else if (request == PTRACE_SYSCALL) {
1da177e4	487	set_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
c8c86cec	488	clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
ed75e8d5	489	} else {
c8c86cec	490	clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
1da177e4 LT	491	clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
	492	}
	493	child->exit_code = data;
	494	/* make sure the single step bit is not set. */
	495	clear_singlestep(child);
	496	wake_up_process(child);
	497	ret = 0;
	498	break;
	499
	500	/*
	501	* make the child exit. Best I can do is send it a sigkill.
	502	* perhaps it should be put in the status that it wants to
	503	* exit.
	504	*/
	505	case PTRACE_KILL:
	506	ret = 0;
	507	if (child->exit_state == EXIT_ZOMBIE) /* already dead */
	508	break;
	509	child->exit_code = SIGKILL;
	510	/* make sure the single step bit is not set. */
	511	clear_singlestep(child);
	512	wake_up_process(child);
	513	break;
	514
1b38f006	515	case PTRACE_SYSEMU_SINGLESTEP: /* Same as SYSEMU, but singlestep if not syscall */
1da177e4 LT	516	case PTRACE_SINGLESTEP: /* set the trap flag. */
1da177e4 LT	517	ret = -EIO;
7ed20e1a	518	if (!valid_signal(data))
1da177e4	519	break;
1b38f006 BS	520
	521	if (request == PTRACE_SYSEMU_SINGLESTEP)
	522	set_tsk_thread_flag(child, TIF_SYSCALL_EMU);
	523	else
	524	clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
	525
1da177e4 LT	526	clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
	527	set_singlestep(child);
	528	child->exit_code = data;
	529	/* give it a chance to run. */
	530	wake_up_process(child);
	531	ret = 0;
	532	break;
	533
	534	case PTRACE_DETACH:
	535	/* detach a process that was attached. */
	536	ret = ptrace_detach(child, data);
	537	break;
	538
	539	case PTRACE_GETREGS: { /* Get all gp regs from the child. */
	540	if (!access_ok(VERIFY_WRITE, datap, FRAME_SIZE*sizeof(long))) {
	541	ret = -EIO;
	542	break;
	543	}
	544	for ( i = 0; i < FRAME_SIZE*sizeof(long); i += sizeof(long) ) {
	545	__put_user(getreg(child, i), datap);
	546	datap++;
	547	}
	548	ret = 0;
	549	break;
	550	}
	551
	552	case PTRACE_SETREGS: { /* Set all gp regs in the child. */
	553	unsigned long tmp;
	554	if (!access_ok(VERIFY_READ, datap, FRAME_SIZE*sizeof(long))) {
	555	ret = -EIO;
	556	break;
	557	}
	558	for ( i = 0; i < FRAME_SIZE*sizeof(long); i += sizeof(long) ) {
	559	__get_user(tmp, datap);
	560	putreg(child, i, tmp);
	561	datap++;
	562	}
	563	ret = 0;
	564	break;
	565	}
	566
	567	case PTRACE_GETFPREGS: { /* Get the child FPU state. */
	568	if (!access_ok(VERIFY_WRITE, datap,
	569	sizeof(struct user_i387_struct))) {
	570	ret = -EIO;
	571	break;
	572	}
	573	ret = 0;
	574	if (!tsk_used_math(child))
	575	init_fpu(child);
	576	get_fpregs((struct user_i387_struct __user *)data, child);
	577	break;
	578	}
	579
	580	case PTRACE_SETFPREGS: { /* Set the child FPU state. */
	581	if (!access_ok(VERIFY_READ, datap,
	582	sizeof(struct user_i387_struct))) {
	583	ret = -EIO;
	584	break;
	585	}
	586	set_stopped_child_used_math(child);
	587	set_fpregs(child, (struct user_i387_struct __user *)data);
	588	ret = 0;
	589	break;
590	}
591
592	case PTRACE_GETFPXREGS: { /* Get the child extended FPU state. */
593	if (!access_ok(VERIFY_WRITE, datap,
594	sizeof(struct user_fxsr_struct))) {
595	ret = -EIO;
596	break;
597	}
598	if (!tsk_used_math(child))
599	init_fpu(child);
600	ret = get_fpxregs((struct user_fxsr_struct __user *)data, child);
601	break;
602	}
603
604	case PTRACE_SETFPXREGS: { /* Set the child extended FPU state. */
605	if (!access_ok(VERIFY_READ, datap,
606	sizeof(struct user_fxsr_struct))) {
607	ret = -EIO;
608	break;
609	}
610	set_stopped_child_used_math(child);
611	ret = set_fpxregs(child, (struct user_fxsr_struct __user *)data);
612	break;
613	}
614
615	case PTRACE_GET_THREAD_AREA:
616	ret = ptrace_get_thread_area(child, addr,
617	(struct user_desc __user *) data);
618	break;
619
620	case PTRACE_SET_THREAD_AREA:
621	ret = ptrace_set_thread_area(child, addr,
622	(struct user_desc __user *) data);
623	break;
624
625	default:
626	ret = ptrace_request(child, request, addr, data);
627	break;
628	}
481bed45	629	out_tsk:
1da177e4 LT	630	return ret;
	631	}
	632
	633	void send_sigtrap(struct task_struct tsk, struct pt_regs regs, int error_code)
	634	{
	635	struct siginfo info;
	636
	637	tsk->thread.trap_no = 1;
	638	tsk->thread.error_code = error_code;
	639
	640	memset(&info, 0, sizeof(info));
	641	info.si_signo = SIGTRAP;
	642	info.si_code = TRAP_BRKPT;
	643
	644	/* User-mode eip? */
fa1e1bdf	645	info.si_addr = user_mode_vm(regs) ? (void __user *) regs->eip : NULL;
1da177e4 LT	646
	647	/* Send us the fakey SIGTRAP */
	648	force_sig_info(SIGTRAP, &info, tsk);
	649	}
	650
	651	/* notification of system call entry/exit
	652	* - triggered by current->work.syscall_trace
	653	*/
	654	__attribute__((regparm(3)))
ed75e8d5	655	int do_syscall_trace(struct pt_regs *regs, int entryexit)
1da177e4	656	{
4c7fc722 AA	657	int is_sysemu = test_thread_flag(TIF_SYSCALL_EMU);
	658	/*
	659	* With TIF_SYSCALL_EMU set we want to ignore TIF_SINGLESTEP for syscall
	660	* interception
	661	*/
1b38f006	662	int is_singlestep = !is_sysemu && test_thread_flag(TIF_SINGLESTEP);
4c7fc722	663	int ret = 0;
1b38f006	664
1da177e4	665	/* do the secure computing check first */
4c7fc722 AA	666	if (!entryexit)
4c7fc722 AA	667	secure_computing(regs->orig_eax);
1da177e4	668
ab1c23c2 BS	669	if (unlikely(current->audit_context)) {
ab1c23c2 BS	670	if (entryexit)
4c7fc722 AA	671	audit_syscall_exit(current, AUDITSC_RESULT(regs->eax),
4c7fc722 AA	672	regs->eax);
ab1c23c2 BS	673	/* Debug traps, when using PTRACE_SINGLESTEP, must be sent only
	674	* on the syscall exit path. Normally, when TIF_SYSCALL_AUDIT is
	675	* not used, entry.S will call us only on syscall exit, not
	676	* entry; so when TIF_SYSCALL_AUDIT is used we must avoid
	677	* calling send_sigtrap() on syscall entry.
	678	*
	679	* Note that when PTRACE_SYSEMU_SINGLESTEP is used,
	680	* is_singlestep is false, despite his name, so we will still do
	681	* the correct thing.
	682	*/
	683	else if (is_singlestep)
	684	goto out;
	685	}
1da177e4 LT	686
1da177e4 LT	687	if (!(current->ptrace & PT_PTRACED))
2fd6f58b	688	goto out;
1da177e4	689
1b38f006 BS	690	/* If a process stops on the 1st tracepoint with SYSCALL_TRACE
	691	* and then is resumed with SYSEMU_SINGLESTEP, it will come in
	692	* here. We have to check this and return */
	693	if (is_sysemu && entryexit)
	694	return 0;
ed75e8d5	695
1da177e4	696	/* Fake a debug trap */
c8c86cec	697	if (is_singlestep)
1da177e4 LT	698	send_sigtrap(current, regs, 0);
1da177e4 LT	699
c8c86cec	700	if (!test_thread_flag(TIF_SYSCALL_TRACE) && !is_sysemu)
2fd6f58b	701	goto out;
1da177e4 LT	702
	703	/* the 0x80 provides a way for the tracing parent to distinguish
	704	between a syscall stop and SIGTRAP delivery */
ed75e8d5	705	/* Note that the debugger could change the result of test_thread_flag!*/
4c7fc722	706	ptrace_notify(SIGTRAP \| ((current->ptrace & PT_TRACESYSGOOD) ? 0x80:0));
1da177e4 LT	707
	708	/*
	709	* this isn't the same as continuing with a signal, but it will do
	710	* for normal use. strace only continues with a signal if the
	711	* stopping signal is not SIGTRAP. -brl
	712	*/
	713	if (current->exit_code) {
	714	send_sig(current->exit_code, current, 1);
	715	current->exit_code = 0;
	716	}
ed75e8d5	717	ret = is_sysemu;
4c7fc722	718	out:
2fd6f58b DW	719	if (unlikely(current->audit_context) && !entryexit)
	720	audit_syscall_entry(current, AUDIT_ARCH_I386, regs->orig_eax,
	721	regs->ebx, regs->ecx, regs->edx, regs->esi);
c8c86cec BS	722	if (ret == 0)
	723	return 0;
	724
1b38f006	725	regs->orig_eax = -1; /* force skip of syscall restarting */
c8c86cec	726	if (unlikely(current->audit_context))
4c7fc722 AA	727	audit_syscall_exit(current, AUDITSC_RESULT(regs->eax),
4c7fc722 AA	728	regs->eax);
c8c86cec	729	return 1;
1da177e4	730	}