[deliverable/linux.git] / arch / x86 / kernel / ptrace_32.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/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].
 * Prohibits changing ID(21), VIP(20), VIF(19), VM(17), NT(14), IOPL(12-13), IF(9).
 * Also masks reserved bits (31-22, 15, 5, 3, 1).
 */
#define FLAG_MASK 0x00050dd5

/*
 * Offset of eflags on child stack..
 */
#define EFL_OFFSET offsetof(struct pt_regs, eflags)

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 bytes into the pt_regs structure on the stack.
 * 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 - sizeof(struct pt_regs);
	stack += offset;
	return (*((int *)stack));
}

/*
 * This routine will put a word on the processes privileged stack.
 * the offset is bytes into the pt_regs structure on the stack.
 * 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 - sizeof(struct pt_regs);
	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 GS:
			if (value && (value & 3) != 3)
				return -EIO;
			child->thread.gs = value;
			return 0;
		case DS:
		case ES:
		case FS:
			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;
			/*
			 * If the user value contains TF, mark that
			 * it was not "us" (the debugger) that set it.
			 * If not, make sure it stays set if we had.
			 */
			if (value & X86_EFLAGS_TF)
				clear_tsk_thread_flag(child, TIF_FORCED_TF);
			else if (test_tsk_thread_flag(child, TIF_FORCED_TF))
				value |= X86_EFLAGS_TF;
			value |= get_stack_long(child, EFL_OFFSET) & ~FLAG_MASK;
			break;
	}
	if (regno > FS*4)
		regno -= 1*4;
	put_stack_long(child, regno, value);
	return 0;
}

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

	switch (regno >> 2) {
		case EFL:
			/*
			 * If the debugger set TF, hide it from the readout.
			 */
			retval = get_stack_long(child, EFL_OFFSET);
			if (test_tsk_thread_flag(child, TIF_FORCED_TF))
				retval &= ~X86_EFLAGS_TF;
			break;
		case GS:
			retval = child->thread.gs;
			break;
		case DS:
		case ES:
		case FS:
		case SS:
		case CS:
			retval = 0xffff;
			/* fall through */
		default:
			if (regno > FS*4)
				regno -= 1*4;
			retval &= get_stack_long(child, regno);
	}
	return retval;
}

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

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:
		ret = generic_ptrace_peekdata(child, addr, data);
		break;

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

		ret = -EIO;
		if ((addr & 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 = generic_ptrace_pokedata(child, addr, data);
		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;
				  if (data)
					  set_tsk_thread_flag(child, TIF_DEBUG);
				  else
					  clear_tsk_thread_flag(child, TIF_DEBUG);
			  }
			  addr -= (long) &dummy->u_debugreg;
			  addr = addr >> 2;
			  child->thread.debugreg[addr] = data;
			  ret = 0;
		  }
		  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:
		if (addr < 0)
			return -EIO;
		ret = do_get_thread_area(child, addr,
					 (struct user_desc __user *) data);
		break;

	case PTRACE_SET_THREAD_AREA:
		if (addr < 0)
			return -EIO;
		ret = do_set_thread_area(child, addr,
					 (struct user_desc __user *) data, 0);
		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 fake 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(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(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(AUDITSC_RESULT(regs->eax), regs->eax);
	return 1;
}
Commit	Line	Data
1da177e4 LT	1	/* By Ross Biro 1/23/92 */
	2	/*
	3	* Pentium III FXSR, SSE support
	4	* Gareth Hughes <gareth@valinux.com>, May 2000
	5	*/
	6
	7	#include <linux/kernel.h>
	8	#include <linux/sched.h>
	9	#include <linux/mm.h>
	10	#include <linux/smp.h>
1da177e4 LT	11	#include <linux/errno.h>
	12	#include <linux/ptrace.h>
	13	#include <linux/user.h>
	14	#include <linux/security.h>
	15	#include <linux/audit.h>
	16	#include <linux/seccomp.h>
7ed20e1a	17	#include <linux/signal.h>
1da177e4 LT	18
	19	#include <asm/uaccess.h>
	20	#include <asm/pgtable.h>
	21	#include <asm/system.h>
	22	#include <asm/processor.h>
	23	#include <asm/i387.h>
	24	#include <asm/debugreg.h>
	25	#include <asm/ldt.h>
	26	#include <asm/desc.h>
	27
	28	/*
	29	* does not yet catch signals sent when the child dies.
	30	* in exit.c or in signal.c.
	31	*/
	32
9f155b98 CE	33	/*
9f155b98 CE	34	* Determines which flags the user has access to [1 = access, 0 = no access].
3c36c6aa	35	* Prohibits changing ID(21), VIP(20), VIF(19), VM(17), NT(14), IOPL(12-13), IF(9).
9f155b98 CE	36	* Also masks reserved bits (31-22, 15, 5, 3, 1).
9f155b98 CE	37	*/
3c36c6aa	38	#define FLAG_MASK 0x00050dd5
1da177e4	39
1da177e4 LT	40	/*
	41	* Offset of eflags on child stack..
	42	*/
8701ea95	43	#define EFL_OFFSET offsetof(struct pt_regs, eflags)
1da177e4 LT	44
	45	static inline struct pt_regs get_child_regs(struct task_struct task)
	46	{
	47	void stack_top = (void )task->thread.esp0;
	48	return stack_top - sizeof(struct pt_regs);
	49	}
	50
	51	/*
8701ea95 JF	52	* This routine will get a word off of the processes privileged stack.
	53	* the offset is bytes into the pt_regs structure on the stack.
	54	* This routine assumes that all the privileged stacks are in our
1da177e4 LT	55	* data space.
	56	*/
	57	static inline int get_stack_long(struct task_struct *task, int offset)
	58	{
	59	unsigned char *stack;
	60
8701ea95	61	stack = (unsigned char *)task->thread.esp0 - sizeof(struct pt_regs);
1da177e4 LT	62	stack += offset;
	63	return (((int )stack));
	64	}
	65
	66	/*
8701ea95 JF	67	* This routine will put a word on the processes privileged stack.
	68	* the offset is bytes into the pt_regs structure on the stack.
	69	* This routine assumes that all the privileged stacks are in our
1da177e4 LT	70	* data space.
	71	*/
	72	static inline int put_stack_long(struct task_struct *task, int offset,
	73	unsigned long data)
	74	{
	75	unsigned char * stack;
	76
8701ea95	77	stack = (unsigned char *)task->thread.esp0 - sizeof(struct pt_regs);
1da177e4 LT	78	stack += offset;
	79	(unsigned long ) stack = data;
	80	return 0;
	81	}
	82
	83	static int putreg(struct task_struct *child,
	84	unsigned long regno, unsigned long value)
	85	{
	86	switch (regno >> 2) {
464d1a78	87	case GS:
1da177e4 LT	88	if (value && (value & 3) != 3)
1da177e4 LT	89	return -EIO;
464d1a78	90	child->thread.gs = value;
1da177e4	91	return 0;
1da177e4 LT	92	case DS:
1da177e4 LT	93	case ES:
464d1a78	94	case FS:
1da177e4 LT	95	if (value && (value & 3) != 3)
	96	return -EIO;
	97	value &= 0xffff;
	98	break;
	99	case SS:
	100	case CS:
	101	if ((value & 3) != 3)
	102	return -EIO;
	103	value &= 0xffff;
	104	break;
	105	case EFL:
	106	value &= FLAG_MASK;
e1f28773 RM	107	/*
	108	* If the user value contains TF, mark that
	109	* it was not "us" (the debugger) that set it.
	110	* If not, make sure it stays set if we had.
	111	*/
	112	if (value & X86_EFLAGS_TF)
	113	clear_tsk_thread_flag(child, TIF_FORCED_TF);
	114	else if (test_tsk_thread_flag(child, TIF_FORCED_TF))
	115	value \|= X86_EFLAGS_TF;
1da177e4 LT	116	value \|= get_stack_long(child, EFL_OFFSET) & ~FLAG_MASK;
	117	break;
	118	}
464d1a78	119	if (regno > FS*4)
66e10a44	120	regno -= 1*4;
8701ea95	121	put_stack_long(child, regno, value);
1da177e4 LT	122	return 0;
	123	}
	124
	125	static unsigned long getreg(struct task_struct *child,
	126	unsigned long regno)
	127	{
	128	unsigned long retval = ~0UL;
	129
	130	switch (regno >> 2) {
e1f28773 RM	131	case EFL:
	132	/*
	133	* If the debugger set TF, hide it from the readout.
	134	*/
	135	retval = get_stack_long(child, EFL_OFFSET);
	136	if (test_tsk_thread_flag(child, TIF_FORCED_TF))
	137	retval &= ~X86_EFLAGS_TF;
	138	break;
464d1a78 JF	139	case GS:
464d1a78 JF	140	retval = child->thread.gs;
1da177e4	141	break;
1da177e4 LT	142	case DS:
1da177e4 LT	143	case ES:
464d1a78	144	case FS:
1da177e4 LT	145	case SS:
	146	case CS:
	147	retval = 0xffff;
	148	/* fall through */
	149	default:
464d1a78	150	if (regno > FS*4)
66e10a44	151	regno -= 1*4;
1da177e4 LT	152	retval &= get_stack_long(child, regno);
	153	}
	154	return retval;
	155	}
	156
1da177e4 LT	157	/*
	158	* Called by kernel/ptrace.c when detaching..
	159	*
	160	* Make sure the single step bit is not set.
	161	*/
	162	void ptrace_disable(struct task_struct *child)
	163	{
7f232343	164	user_disable_single_step(child);
ab1c23c2	165	clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
1da177e4 LT	166	}
1da177e4 LT	167
481bed45	168	long arch_ptrace(struct task_struct *child, long request, long addr, long data)
1da177e4	169	{
1da177e4 LT	170	struct user * dummy = NULL;
	171	int i, ret;
	172	unsigned long __user datap = (unsigned long __user )data;
	173
1da177e4 LT	174	switch (request) {
	175	/* when I and D space are separate, these will need to be fixed. */
	176	case PTRACE_PEEKTEXT: /* read word at location addr. */
76647323 AD	177	case PTRACE_PEEKDATA:
76647323 AD	178	ret = generic_ptrace_peekdata(child, addr, data);
1da177e4	179	break;
1da177e4 LT	180
	181	/* read the word at location addr in the USER area. */
	182	case PTRACE_PEEKUSR: {
	183	unsigned long tmp;
	184
	185	ret = -EIO;
	186	if ((addr & 3) \|\| addr < 0 \|\|
	187	addr > sizeof(struct user) - 3)
	188	break;
	189
	190	tmp = 0; /* Default return condition */
	191	if(addr < FRAME_SIZE*sizeof(long))
	192	tmp = getreg(child, addr);
	193	if(addr >= (long) &dummy->u_debugreg[0] &&
	194	addr <= (long) &dummy->u_debugreg[7]){
	195	addr -= (long) &dummy->u_debugreg[0];
	196	addr = addr >> 2;
	197	tmp = child->thread.debugreg[addr];
	198	}
	199	ret = put_user(tmp, datap);
	200	break;
	201	}
	202
	203	/* when I and D space are separate, this will have to be fixed. */
	204	case PTRACE_POKETEXT: /* write the word at location addr. */
	205	case PTRACE_POKEDATA:
f284ce72	206	ret = generic_ptrace_pokedata(child, addr, data);
1da177e4 LT	207	break;
	208
	209	case PTRACE_POKEUSR: /* write the word at location addr in the USER area */
	210	ret = -EIO;
	211	if ((addr & 3) \|\| addr < 0 \|\|
	212	addr > sizeof(struct user) - 3)
	213	break;
	214
	215	if (addr < FRAME_SIZE*sizeof(long)) {
	216	ret = putreg(child, addr, data);
	217	break;
	218	}
	219	/* We need to be very careful here. We implicitly
	220	want to modify a portion of the task_struct, and we
	221	have to be selective about what portions we allow someone
	222	to modify. */
	223
	224	ret = -EIO;
	225	if(addr >= (long) &dummy->u_debugreg[0] &&
	226	addr <= (long) &dummy->u_debugreg[7]){
	227
	228	if(addr == (long) &dummy->u_debugreg[4]) break;
	229	if(addr == (long) &dummy->u_debugreg[5]) break;
	230	if(addr < (long) &dummy->u_debugreg[4] &&
	231	((unsigned long) data) >= TASK_SIZE-3) break;
	232
	233	/* Sanity-check data. Take one half-byte at once with
	234	* check = (val >> (16 + 4*i)) & 0xf. It contains the
	235	* R/Wi and LENi bits; bits 0 and 1 are R/Wi, and bits
	236	* 2 and 3 are LENi. Given a list of invalid values,
	237	* we do mask \|= 1 << invalid_value, so that
	238	* (mask >> check) & 1 is a correct test for invalid
	239	* values.
	240	*
	241	* R/Wi contains the type of the breakpoint /
	242	* watchpoint, LENi contains the length of the watched
	243	* data in the watchpoint case.
	244	*
	245	* The invalid values are:
	246	* - LENi == 0x10 (undefined), so mask \|= 0x0f00.
	247	* - R/Wi == 0x10 (break on I/O reads or writes), so
	248	* mask \|= 0x4444.
	249	* - R/Wi == 0x00 && LENi != 0x00, so we have mask \|=
	250	* 0x1110.
	251	*
	252	* Finally, mask = 0x0f00 \| 0x4444 \| 0x1110 == 0x5f54.
	253	*
	254	* See the Intel Manual "System Programming Guide",
	255	* 15.2.4
	256	*
	257	* Note that LENi == 0x10 is defined on x86_64 in long
	258	* mode (i.e. even for 32-bit userspace software, but
	259	* 64-bit kernel), so the x86_64 mask value is 0x5454.
	260	* See the AMD manual no. 24593 (AMD64 System
	261	* Programming)*/
	262
	263	if(addr == (long) &dummy->u_debugreg[7]) {
	264	data &= ~DR_CONTROL_RESERVED;
	265	for(i=0; i<4; i++)
	266	if ((0x5f54 >> ((data >> (16 + 4*i)) & 0xf)) & 1)
	267	goto out_tsk;
b3cf2576 SE	268	if (data)
	269	set_tsk_thread_flag(child, TIF_DEBUG);
	270	else
	271	clear_tsk_thread_flag(child, TIF_DEBUG);
1da177e4	272	}
1da177e4 LT	273	addr -= (long) &dummy->u_debugreg;
	274	addr = addr >> 2;
	275	child->thread.debugreg[addr] = data;
	276	ret = 0;
	277	}
	278	break;
	279
1da177e4 LT	280	case PTRACE_GETREGS: { /* Get all gp regs from the child. */
	281	if (!access_ok(VERIFY_WRITE, datap, FRAME_SIZE*sizeof(long))) {
	282	ret = -EIO;
	283	break;
	284	}
	285	for ( i = 0; i < FRAME_SIZE*sizeof(long); i += sizeof(long) ) {
	286	__put_user(getreg(child, i), datap);
	287	datap++;
	288	}
	289	ret = 0;
	290	break;
	291	}
	292
	293	case PTRACE_SETREGS: { /* Set all gp regs in the child. */
	294	unsigned long tmp;
	295	if (!access_ok(VERIFY_READ, datap, FRAME_SIZE*sizeof(long))) {
	296	ret = -EIO;
	297	break;
	298	}
	299	for ( i = 0; i < FRAME_SIZE*sizeof(long); i += sizeof(long) ) {
	300	__get_user(tmp, datap);
	301	putreg(child, i, tmp);
	302	datap++;
	303	}
	304	ret = 0;
	305	break;
	306	}
	307
	308	case PTRACE_GETFPREGS: { /* Get the child FPU state. */
	309	if (!access_ok(VERIFY_WRITE, datap,
	310	sizeof(struct user_i387_struct))) {
	311	ret = -EIO;
	312	break;
	313	}
	314	ret = 0;
	315	if (!tsk_used_math(child))
	316	init_fpu(child);
	317	get_fpregs((struct user_i387_struct __user *)data, child);
	318	break;
	319	}
	320
	321	case PTRACE_SETFPREGS: { /* Set the child FPU state. */
	322	if (!access_ok(VERIFY_READ, datap,
	323	sizeof(struct user_i387_struct))) {
	324	ret = -EIO;
	325	break;
	326	}
	327	set_stopped_child_used_math(child);
	328	set_fpregs(child, (struct user_i387_struct __user *)data);
	329	ret = 0;
	330	break;
	331	}
	332
	333	case PTRACE_GETFPXREGS: { /* Get the child extended FPU state. */
	334	if (!access_ok(VERIFY_WRITE, datap,
	335	sizeof(struct user_fxsr_struct))) {
	336	ret = -EIO;
	337	break;
	338	}
	339	if (!tsk_used_math(child))
	340	init_fpu(child);
	341	ret = get_fpxregs((struct user_fxsr_struct __user *)data, child);
	342	break;
	343	}
344
345	case PTRACE_SETFPXREGS: { /* Set the child extended FPU state. */
346	if (!access_ok(VERIFY_READ, datap,
347	sizeof(struct user_fxsr_struct))) {
348	ret = -EIO;
349	break;
350	}
351	set_stopped_child_used_math(child);
352	ret = set_fpxregs(child, (struct user_fxsr_struct __user *)data);
353	break;
354	}
355
356	case PTRACE_GET_THREAD_AREA:
efd1ca52 RM	357	if (addr < 0)
	358	return -EIO;
	359	ret = do_get_thread_area(child, addr,
	360	(struct user_desc __user *) data);
1da177e4 LT	361	break;
	362
	363	case PTRACE_SET_THREAD_AREA:
efd1ca52 RM	364	if (addr < 0)
	365	return -EIO;
	366	ret = do_set_thread_area(child, addr,
	367	(struct user_desc __user *) data, 0);
1da177e4 LT	368	break;
	369
	370	default:
	371	ret = ptrace_request(child, request, addr, data);
	372	break;
	373	}
481bed45	374	out_tsk:
1da177e4 LT	375	return ret;
	376	}
	377
	378	void send_sigtrap(struct task_struct tsk, struct pt_regs regs, int error_code)
	379	{
	380	struct siginfo info;
	381
	382	tsk->thread.trap_no = 1;
	383	tsk->thread.error_code = error_code;
	384
	385	memset(&info, 0, sizeof(info));
	386	info.si_signo = SIGTRAP;
	387	info.si_code = TRAP_BRKPT;
	388
	389	/* User-mode eip? */
fa1e1bdf	390	info.si_addr = user_mode_vm(regs) ? (void __user *) regs->eip : NULL;
1da177e4	391
27b46d76	392	/* Send us the fake SIGTRAP */
1da177e4 LT	393	force_sig_info(SIGTRAP, &info, tsk);
	394	}
	395
	396	/* notification of system call entry/exit
	397	* - triggered by current->work.syscall_trace
	398	*/
	399	__attribute__((regparm(3)))
ed75e8d5	400	int do_syscall_trace(struct pt_regs *regs, int entryexit)
1da177e4	401	{
4c7fc722 AA	402	int is_sysemu = test_thread_flag(TIF_SYSCALL_EMU);
	403	/*
	404	* With TIF_SYSCALL_EMU set we want to ignore TIF_SINGLESTEP for syscall
	405	* interception
	406	*/
1b38f006	407	int is_singlestep = !is_sysemu && test_thread_flag(TIF_SINGLESTEP);
4c7fc722	408	int ret = 0;
1b38f006	409
1da177e4	410	/* do the secure computing check first */
4c7fc722 AA	411	if (!entryexit)
4c7fc722 AA	412	secure_computing(regs->orig_eax);
1da177e4	413
ab1c23c2 BS	414	if (unlikely(current->audit_context)) {
ab1c23c2 BS	415	if (entryexit)
5411be59	416	audit_syscall_exit(AUDITSC_RESULT(regs->eax),
4c7fc722	417	regs->eax);
ab1c23c2 BS	418	/* Debug traps, when using PTRACE_SINGLESTEP, must be sent only
	419	* on the syscall exit path. Normally, when TIF_SYSCALL_AUDIT is
	420	* not used, entry.S will call us only on syscall exit, not
	421	* entry; so when TIF_SYSCALL_AUDIT is used we must avoid
	422	* calling send_sigtrap() on syscall entry.
	423	*
	424	* Note that when PTRACE_SYSEMU_SINGLESTEP is used,
	425	* is_singlestep is false, despite his name, so we will still do
	426	* the correct thing.
	427	*/
	428	else if (is_singlestep)
	429	goto out;
	430	}
1da177e4 LT	431
1da177e4 LT	432	if (!(current->ptrace & PT_PTRACED))
2fd6f58b	433	goto out;
1da177e4	434
1b38f006 BS	435	/* If a process stops on the 1st tracepoint with SYSCALL_TRACE
	436	* and then is resumed with SYSEMU_SINGLESTEP, it will come in
	437	* here. We have to check this and return */
	438	if (is_sysemu && entryexit)
	439	return 0;
ed75e8d5	440
1da177e4	441	/* Fake a debug trap */
c8c86cec	442	if (is_singlestep)
1da177e4 LT	443	send_sigtrap(current, regs, 0);
1da177e4 LT	444
c8c86cec	445	if (!test_thread_flag(TIF_SYSCALL_TRACE) && !is_sysemu)
2fd6f58b	446	goto out;
1da177e4 LT	447
	448	/* the 0x80 provides a way for the tracing parent to distinguish
	449	between a syscall stop and SIGTRAP delivery */
ed75e8d5	450	/* Note that the debugger could change the result of test_thread_flag!*/
4c7fc722	451	ptrace_notify(SIGTRAP \| ((current->ptrace & PT_TRACESYSGOOD) ? 0x80:0));
1da177e4 LT	452
	453	/*
	454	* this isn't the same as continuing with a signal, but it will do
	455	* for normal use. strace only continues with a signal if the
	456	* stopping signal is not SIGTRAP. -brl
	457	*/
	458	if (current->exit_code) {
	459	send_sig(current->exit_code, current, 1);
	460	current->exit_code = 0;
	461	}
ed75e8d5	462	ret = is_sysemu;
4c7fc722	463	out:
2fd6f58b	464	if (unlikely(current->audit_context) && !entryexit)
5411be59	465	audit_syscall_entry(AUDIT_ARCH_I386, regs->orig_eax,
2fd6f58b	466	regs->ebx, regs->ecx, regs->edx, regs->esi);
c8c86cec BS	467	if (ret == 0)
	468	return 0;
	469
1b38f006	470	regs->orig_eax = -1; /* force skip of syscall restarting */
c8c86cec	471	if (unlikely(current->audit_context))
5411be59	472	audit_syscall_exit(AUDITSC_RESULT(regs->eax), regs->eax);
c8c86cec	473	return 1;
1da177e4	474	}