Pull utrace into release branch

arch/ia64/Kconfig

@@ -21,6 +21,7 @@ config IA64
 	select HAVE_KRETPROBES
 	select HAVE_DMA_ATTRS
 	select HAVE_KVM
+	select HAVE_ARCH_TRACEHOOK
 	default y
 	help
 	  The Itanium Processor Family is Intel's 64-bit successor to

arch/ia64/ia32/ia32_entry.S

@@ -108,6 +108,11 @@ GLOBAL_ENTRY(ia32_trace_syscall)
 	;;
 	st8 [r2]=r3				// initialize return code to -ENOSYS
 	br.call.sptk.few rp=syscall_trace_enter	// give parent a chance to catch syscall args
+	cmp.lt p6,p0=r8,r0			// check tracehook
+	adds r2=IA64_PT_REGS_R8_OFFSET+16,sp	// r2 = &pt_regs.r8
+	;;
+(p6)	st8.spill [r2]=r8			// store return value in slot for r8
+(p6)	br.spnt.few .ret4
 .ret2:	// Need to reload arguments (they may be changed by the tracing process)
 	adds r2=IA64_PT_REGS_R1_OFFSET+16,sp	// r2 = &pt_regs.r1
 	adds r3=IA64_PT_REGS_R13_OFFSET+16,sp	// r3 = &pt_regs.r13
@@ -199,7 +204,7 @@ ia32_syscall_table:
 	data8 sys_setuid	/* 16-bit version */
 	data8 sys_getuid	/* 16-bit version */
 	data8 compat_sys_stime    /* 25 */
-	data8 sys32_ptrace
+	data8 compat_sys_ptrace
 	data8 sys32_alarm
 	data8 sys_ni_syscall
 	data8 sys_pause

arch/ia64/ia32/sys_ia32.c

@@ -1194,25 +1194,6 @@ sys32_waitpid (int pid, unsigned int *stat_addr, int options)
 	return compat_sys_wait4(pid, stat_addr, options, NULL);
 }
 
-static unsigned int
-ia32_peek (struct task_struct *child, unsigned long addr, unsigned int *val)
-{
-	size_t copied;
-	unsigned int ret;
-
-	copied = access_process_vm(child, addr, val, sizeof(*val), 0);
-	return (copied != sizeof(ret)) ? -EIO : 0;
-}
-
-static unsigned int
-ia32_poke (struct task_struct *child, unsigned long addr, unsigned int val)
-{
-
-	if (access_process_vm(child, addr, &val, sizeof(val), 1) != sizeof(val))
-		return -EIO;
-	return 0;
-}
-
 /*
  *  The order in which registers are stored in the ptrace regs structure
  */
@@ -1510,49 +1491,15 @@ restore_ia32_fpxstate (struct task_struct *tsk, struct ia32_user_fxsr_struct __u
 	return 0;
 }
 
-asmlinkage long
-sys32_ptrace (int request, pid_t pid, unsigned int addr, unsigned int data)
+long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
+	compat_ulong_t caddr, compat_ulong_t cdata)
 {
-	struct task_struct *child;
-	unsigned int value, tmp;
+	unsigned long addr = caddr;
+	unsigned long data = cdata;
+	unsigned int tmp;
 	long i, ret;
 
-	lock_kernel();
-	if (request == PTRACE_TRACEME) {
-		ret = ptrace_traceme();
-		goto out;
-	}
-
-	child = ptrace_get_task_struct(pid);
-	if (IS_ERR(child)) {
-		ret = PTR_ERR(child);
-		goto out;
-	}
-
-	if (request == PTRACE_ATTACH) {
-		ret = sys_ptrace(request, pid, addr, data);
-		goto out_tsk;
-	}
-
-	ret = ptrace_check_attach(child, request == PTRACE_KILL);
-	if (ret < 0)
-		goto out_tsk;
-
 	switch (request) {
-	      case PTRACE_PEEKTEXT:
-	      case PTRACE_PEEKDATA:	/* read word at location addr */
-		ret = ia32_peek(child, addr, &value);
-		if (ret == 0)
-			ret = put_user(value, (unsigned int __user *) compat_ptr(data));
-		else
-			ret = -EIO;
-		goto out_tsk;
-
-	      case PTRACE_POKETEXT:
-	      case PTRACE_POKEDATA:	/* write the word at location addr */
-		ret = ia32_poke(child, addr, data);
-		goto out_tsk;
-
 	      case PTRACE_PEEKUSR:	/* read word at addr in USER area */
 		ret = -EIO;
 		if ((addr & 3) || addr > 17*sizeof(int))
@@ -1617,27 +1564,9 @@ sys32_ptrace (int request, pid_t pid, unsigned int addr, unsigned int data)
 					    compat_ptr(data));
 		break;
 
-	      case PTRACE_GETEVENTMSG:   
-		ret = put_user(child->ptrace_message, (unsigned int __user *) compat_ptr(data));
-		break;
-
-	      case PTRACE_SYSCALL:	/* continue, stop after next syscall */
-	      case PTRACE_CONT:		/* restart after signal. */
-	      case PTRACE_KILL:
-	      case PTRACE_SINGLESTEP:	/* execute chile for one instruction */
-	      case PTRACE_DETACH:	/* detach a process */
-		ret = sys_ptrace(request, pid, addr, data);
-		break;
-
 	      default:
-		ret = ptrace_request(child, request, addr, data);
-		break;
-
+		return compat_ptrace_request(child, request, caddr, cdata);
 	}
-  out_tsk:
-	put_task_struct(child);
-  out:
-	unlock_kernel();
 	return ret;
 }
 

arch/ia64/include/asm/ptrace.h

@@ -240,6 +240,12 @@ struct switch_stack {
  */
 # define instruction_pointer(regs) ((regs)->cr_iip + ia64_psr(regs)->ri)
 
+static inline unsigned long user_stack_pointer(struct pt_regs *regs)
+{
+	/* FIXME: should this be bspstore + nr_dirty regs? */
+	return regs->ar_bspstore;
+}
+
 #define regs_return_value(regs) ((regs)->r8)
 
 /* Conserve space in histogram by encoding slot bits in address
@@ -319,6 +325,8 @@ struct switch_stack {
   #define arch_has_block_step()   (1)
   extern void user_enable_block_step(struct task_struct *);
 
+#define __ARCH_WANT_COMPAT_SYS_PTRACE
+
 #endif /* !__KERNEL__ */
 
 /* pt_all_user_regs is used for PTRACE_GETREGS PTRACE_SETREGS */

arch/ia64/include/asm/syscall.h

@@ -0,0 +1,163 @@
+/*
+ * Access to user system call parameters and results
+ *
+ * Copyright (C) 2008 Intel Corp.  Shaohua Li <shaohua.li@intel.com>
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU General Public License v.2.
+ *
+ * See asm-generic/syscall.h for descriptions of what we must do here.
+ */
+
+#ifndef _ASM_SYSCALL_H
+#define _ASM_SYSCALL_H	1
+
+#include <linux/sched.h>
+#include <linux/err.h>
+
+static inline long syscall_get_nr(struct task_struct *task,
+				  struct pt_regs *regs)
+{
+	if ((long)regs->cr_ifs < 0) /* Not a syscall */
+		return -1;
+
+#ifdef CONFIG_IA32_SUPPORT
+	if (IS_IA32_PROCESS(regs))
+		return regs->r1;
+#endif
+
+	return regs->r15;
+}
+
+static inline void syscall_rollback(struct task_struct *task,
+				    struct pt_regs *regs)
+{
+#ifdef CONFIG_IA32_SUPPORT
+	if (IS_IA32_PROCESS(regs))
+		regs->r8 = regs->r1;
+#endif
+
+	/* do nothing */
+}
+
+static inline long syscall_get_error(struct task_struct *task,
+				     struct pt_regs *regs)
+{
+#ifdef CONFIG_IA32_SUPPORT
+	if (IS_IA32_PROCESS(regs))
+		return regs->r8;
+#endif
+
+	return regs->r10 == -1 ? regs->r8:0;
+}
+
+static inline long syscall_get_return_value(struct task_struct *task,
+					    struct pt_regs *regs)
+{
+	return regs->r8;
+}
+
+static inline void syscall_set_return_value(struct task_struct *task,
+					    struct pt_regs *regs,
+					    int error, long val)
+{
+#ifdef CONFIG_IA32_SUPPORT
+	if (IS_IA32_PROCESS(regs)) {
+		regs->r8 = (long) error ? error : val;
+		return;
+	}
+#endif
+
+	if (error) {
+		/* error < 0, but ia64 uses > 0 return value */
+		regs->r8 = -error;
+		regs->r10 = -1;
+	} else {
+		regs->r8 = val;
+		regs->r10 = 0;
+	}
+}
+
+extern void ia64_syscall_get_set_arguments(struct task_struct *task,
+	struct pt_regs *regs, unsigned int i, unsigned int n,
+	unsigned long *args, int rw);
+static inline void syscall_get_arguments(struct task_struct *task,
+					 struct pt_regs *regs,
+					 unsigned int i, unsigned int n,
+					 unsigned long *args)
+{
+	BUG_ON(i + n > 6);
+
+#ifdef CONFIG_IA32_SUPPORT
+	if (IS_IA32_PROCESS(regs)) {
+		switch (i + n) {
+		case 6:
+			if (!n--) break;
+			*args++ = regs->r13;
+		case 5:
+			if (!n--) break;
+			*args++ = regs->r15;
+		case 4:
+			if (!n--) break;
+			*args++ = regs->r14;
+		case 3:
+			if (!n--) break;
+			*args++ = regs->r10;
+		case 2:
+			if (!n--) break;
+			*args++ = regs->r9;
+		case 1:
+			if (!n--) break;
+			*args++ = regs->r11;
+		case 0:
+			if (!n--) break;
+		default:
+			BUG();
+			break;
+		}
+
+		return;
+	}
+#endif
+	ia64_syscall_get_set_arguments(task, regs, i, n, args, 0);
+}
+
+static inline void syscall_set_arguments(struct task_struct *task,
+					 struct pt_regs *regs,
+					 unsigned int i, unsigned int n,
+					 unsigned long *args)
+{
+	BUG_ON(i + n > 6);
+
+#ifdef CONFIG_IA32_SUPPORT
+	if (IS_IA32_PROCESS(regs)) {
+		switch (i + n) {
+		case 6:
+			if (!n--) break;
+			regs->r13 = *args++;
+		case 5:
+			if (!n--) break;
+			regs->r15 = *args++;
+		case 4:
+			if (!n--) break;
+			regs->r14 = *args++;
+		case 3:
+			if (!n--) break;
+			regs->r10 = *args++;
+		case 2:
+			if (!n--) break;
+			regs->r9 = *args++;
+		case 1:
+			if (!n--) break;
+			regs->r11 = *args++;
+		case 0:
+			if (!n--) break;
+		}
+
+		return;
+	}
+#endif
+	ia64_syscall_get_set_arguments(task, regs, i, n, args, 1);
+}
+#endif	/* _ASM_SYSCALL_H */

arch/ia64/include/asm/thread_info.h

@@ -87,9 +87,6 @@ struct thread_info {
 #define alloc_task_struct()	((struct task_struct *)__get_free_pages(GFP_KERNEL | __GFP_COMP, KERNEL_STACK_SIZE_ORDER))
 #define free_task_struct(tsk)	free_pages((unsigned long) (tsk), KERNEL_STACK_SIZE_ORDER)
 
-#define tsk_set_notify_resume(tsk) \
-	set_ti_thread_flag(task_thread_info(tsk), TIF_NOTIFY_RESUME)
-extern void tsk_clear_notify_resume(struct task_struct *tsk);
 #endif /* !__ASSEMBLY */
 
 /*

arch/ia64/kernel/entry.S

@@ -534,6 +534,11 @@ GLOBAL_ENTRY(ia64_trace_syscall)
  	stf.spill [r16]=f10
  	stf.spill [r17]=f11
 	br.call.sptk.many rp=syscall_trace_enter // give parent a chance to catch syscall args
+	cmp.lt p6,p0=r8,r0			// check tracehook
+	adds r2=PT(R8)+16,sp			// r2 = &pt_regs.r8
+	adds r3=PT(R10)+16,sp			// r3 = &pt_regs.r10
+	mov r10=0
+(p6)	br.cond.sptk strace_error		// syscall failed ->
 	adds r16=PT(F6)+16,sp
 	adds r17=PT(F7)+16,sp
 	;;

arch/ia64/kernel/perfmon.c

@@ -40,6 +40,7 @@
 #include <linux/capability.h>
 #include <linux/rcupdate.h>
 #include <linux/completion.h>
+#include <linux/tracehook.h>
 
 #include <asm/errno.h>
 #include <asm/intrinsics.h>
@@ -3684,7 +3685,7 @@ pfm_restart(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs)
 
 		PFM_SET_WORK_PENDING(task, 1);
 
-		tsk_set_notify_resume(task);
+		set_notify_resume(task);
 
 		/*
 		 * XXX: send reschedule if task runs on another CPU
@@ -5044,8 +5045,6 @@ pfm_handle_work(void)
 
 	PFM_SET_WORK_PENDING(current, 0);
 
-	tsk_clear_notify_resume(current);
-
 	regs = task_pt_regs(current);
 
 	/*
@@ -5414,7 +5413,7 @@ pfm_overflow_handler(struct task_struct *task, pfm_context_t *ctx, u64 pmc0, str
 			 * when coming from ctxsw, current still points to the
 			 * previous task, therefore we must work with task and not current.
 			 */
-			tsk_set_notify_resume(task);
+			set_notify_resume(task);
 		}
 		/*
 		 * defer until state is changed (shorten spin window). the context is locked

arch/ia64/kernel/process.c

@@ -28,6 +28,7 @@
 #include <linux/delay.h>
 #include <linux/kdebug.h>
 #include <linux/utsname.h>
+#include <linux/tracehook.h>
 
 #include <asm/cpu.h>
 #include <asm/delay.h>
@@ -160,21 +161,6 @@ show_regs (struct pt_regs *regs)
 		show_stack(NULL, NULL);
 }
 
-void tsk_clear_notify_resume(struct task_struct *tsk)
-{
-#ifdef CONFIG_PERFMON
-	if (tsk->thread.pfm_needs_checking)
-		return;
-#endif
-	if (test_ti_thread_flag(task_thread_info(tsk), TIF_RESTORE_RSE))
-		return;
-	clear_ti_thread_flag(task_thread_info(tsk), TIF_NOTIFY_RESUME);
-}
-
-/*
- * do_notify_resume_user():
- *	Called from notify_resume_user at entry.S, with interrupts disabled.
- */
 void
 do_notify_resume_user(sigset_t *unused, struct sigscratch *scr, long in_syscall)
 {
@@ -203,6 +189,11 @@ do_notify_resume_user(sigset_t *unused, struct sigscratch *scr, long in_syscall)
 		ia64_do_signal(scr, in_syscall);
 	}
 
+	if (test_thread_flag(TIF_NOTIFY_RESUME)) {
+		clear_thread_flag(TIF_NOTIFY_RESUME);
+		tracehook_notify_resume(&scr->pt);
+	}
+
 	/* copy user rbs to kernel rbs */
 	if (unlikely(test_thread_flag(TIF_RESTORE_RSE))) {
 		local_irq_enable();	/* force interrupt enable */

arch/ia64/kernel/ptrace.c

@@ -22,6 +22,7 @@
 #include <linux/signal.h>
 #include <linux/regset.h>
 #include <linux/elf.h>
+#include <linux/tracehook.h>
 
 #include <asm/pgtable.h>
 #include <asm/processor.h>
@@ -603,7 +604,7 @@ void ia64_ptrace_stop(void)
 {
 	if (test_and_set_tsk_thread_flag(current, TIF_RESTORE_RSE))
 		return;
-	tsk_set_notify_resume(current);
+	set_notify_resume(current);
 	unw_init_running(do_sync_rbs, ia64_sync_user_rbs);
 }
 
@@ -613,7 +614,6 @@ void ia64_ptrace_stop(void)
 void ia64_sync_krbs(void)
 {
 	clear_tsk_thread_flag(current, TIF_RESTORE_RSE);
-	tsk_clear_notify_resume(current);
 
 	unw_init_running(do_sync_rbs, ia64_sync_kernel_rbs);
 }
@@ -644,7 +644,7 @@ ptrace_attach_sync_user_rbs (struct task_struct *child)
 		spin_lock_irq(&child->sighand->siglock);
 		if (child->state == TASK_STOPPED &&
 		    !test_and_set_tsk_thread_flag(child, TIF_RESTORE_RSE)) {
-			tsk_set_notify_resume(child);
+			set_notify_resume(child);
 
 			child->state = TASK_TRACED;
 			stopped = 1;
@@ -1232,37 +1232,16 @@ arch_ptrace (struct task_struct *child, long request, long addr, long data)
 }
 
 
-static void
-syscall_trace (void)
-{
-	/*
-	 * The 0x80 provides a way for the tracing parent to
-	 * distinguish between a syscall stop and SIGTRAP delivery.
-	 */
-	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;
-	}
-}
-
 /* "asmlinkage" so the input arguments are preserved... */
 
-asmlinkage void
+asmlinkage long
 syscall_trace_enter (long arg0, long arg1, long arg2, long arg3,
 		     long arg4, long arg5, long arg6, long arg7,
 		     struct pt_regs regs)
 {
-	if (test_thread_flag(TIF_SYSCALL_TRACE) 
-	    && (current->ptrace & PT_PTRACED))
-		syscall_trace();
+	if (test_thread_flag(TIF_SYSCALL_TRACE))
+		if (tracehook_report_syscall_entry(&regs))
+			return -ENOSYS;
 
 	/* copy user rbs to kernel rbs */
 	if (test_thread_flag(TIF_RESTORE_RSE))
@@ -1283,6 +1262,7 @@ syscall_trace_enter (long arg0, long arg1, long arg2, long arg3,
 		audit_syscall_entry(arch, syscall, arg0, arg1, arg2, arg3);
 	}
 
+	return 0;
 }
 
 /* "asmlinkage" so the input arguments are preserved... */
@@ -1292,6 +1272,8 @@ syscall_trace_leave (long arg0, long arg1, long arg2, long arg3,
 		     long arg4, long arg5, long arg6, long arg7,
 		     struct pt_regs regs)
 {
+	int step;
+
 	if (unlikely(current->audit_context)) {
 		int success = AUDITSC_RESULT(regs.r10);
 		long result = regs.r8;
@@ -1301,10 +1283,9 @@ syscall_trace_leave (long arg0, long arg1, long arg2, long arg3,
 		audit_syscall_exit(success, result);
 	}
 
-	if ((test_thread_flag(TIF_SYSCALL_TRACE)
-	    || test_thread_flag(TIF_SINGLESTEP))
-	    && (current->ptrace & PT_PTRACED))
-		syscall_trace();
+	step = test_thread_flag(TIF_SINGLESTEP);
+	if (step || test_thread_flag(TIF_SYSCALL_TRACE))
+		tracehook_report_syscall_exit(&regs, step);
 
 	/* copy user rbs to kernel rbs */
 	if (test_thread_flag(TIF_RESTORE_RSE))
@@ -1940,7 +1921,7 @@ gpregs_writeback(struct task_struct *target,
 {
 	if (test_and_set_tsk_thread_flag(target, TIF_RESTORE_RSE))
 		return 0;
-	tsk_set_notify_resume(target);
+	set_notify_resume(target);
 	return do_regset_call(do_gpregs_writeback, target, regset, 0, 0,
 		NULL, NULL);
 }
@@ -2199,3 +2180,68 @@ const struct user_regset_view *task_user_regset_view(struct task_struct *tsk)
 #endif
 	return &user_ia64_view;
 }
+
+struct syscall_get_set_args {
+	unsigned int i;
+	unsigned int n;
+	unsigned long *args;
+	struct pt_regs *regs;
+	int rw;
+};
+
+static void syscall_get_set_args_cb(struct unw_frame_info *info, void *data)
+{
+	struct syscall_get_set_args *args = data;
+	struct pt_regs *pt = args->regs;
+	unsigned long *krbs, cfm, ndirty;
+	int i, count;
+
+	if (unw_unwind_to_user(info) < 0)
+		return;
+
+	cfm = pt->cr_ifs;
+	krbs = (unsigned long *)info->task + IA64_RBS_OFFSET/8;
+	ndirty = ia64_rse_num_regs(krbs, krbs + (pt->loadrs >> 19));
+
+	count = 0;
+	if (in_syscall(pt))
+		count = min_t(int, args->n, cfm & 0x7f);
+
+	for (i = 0; i < count; i++) {
+		if (args->rw)
+			*ia64_rse_skip_regs(krbs, ndirty + i + args->i) =
+				args->args[i];
+		else
+			args->args[i] = *ia64_rse_skip_regs(krbs,
+				ndirty + i + args->i);
+	}
+
+	if (!args->rw) {
+		while (i < args->n) {
+			args->args[i] = 0;
+			i++;
+		}
+	}
+}
+
+void ia64_syscall_get_set_arguments(struct task_struct *task,
+	struct pt_regs *regs, unsigned int i, unsigned int n,
+	unsigned long *args, int rw)
+{
+	struct syscall_get_set_args data = {
+		.i = i,
+		.n = n,
+		.args = args,
+		.regs = regs,
+		.rw = rw,
+	};
+
+	if (task == current)
+		unw_init_running(syscall_get_set_args_cb, &data);
+	else {
+		struct unw_frame_info ufi;
+		memset(&ufi, 0, sizeof(ufi));
+		unw_init_from_blocked_task(&ufi, task);
+		syscall_get_set_args_cb(&ufi, &data);
+	}
+}

arch/ia64/kernel/signal.c

@@ -11,6 +11,7 @@
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/ptrace.h>
+#include <linux/tracehook.h>
 #include <linux/sched.h>
 #include <linux/signal.h>
 #include <linux/smp.h>
@@ -439,6 +440,13 @@ handle_signal (unsigned long sig, struct k_sigaction *ka, siginfo_t *info, sigse
 		sigaddset(&current->blocked, sig);
 	recalc_sigpending();
 	spin_unlock_irq(&current->sighand->siglock);
+
+	/*
+	 * Let tracing know that we've done the handler setup.
+	 */
+	tracehook_signal_handler(sig, info, ka, &scr->pt,
+				 test_thread_flag(TIF_SINGLESTEP));
+
 	return 1;
 }