Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/signal

Pull big execve/kernel_thread/fork unification series from Al Viro:
 "All architectures are converted to new model.  Quite a bit of that
  stuff is actually shared with architecture trees; in such cases it's
  literally shared branch pulled by both, not a cherry-pick.

  A lot of ugliness and black magic is gone (-3KLoC total in this one):

   - kernel_thread()/kernel_execve()/sys_execve() redesign.

     We don't do syscalls from kernel anymore for either kernel_thread()
     or kernel_execve():

     kernel_thread() is essentially clone(2) with callback run before we
     return to userland, the callbacks either never return or do
     successful do_execve() before returning.

     kernel_execve() is a wrapper for do_execve() - it doesn't need to
     do transition to user mode anymore.

     As a result kernel_thread() and kernel_execve() are
     arch-independent now - they live in kernel/fork.c and fs/exec.c
     resp.  sys_execve() is also in fs/exec.c and it's completely
     architecture-independent.

   - daemonize() is gone, along with its parts in fs/*.c

   - struct pt_regs * is no longer passed to do_fork/copy_process/
     copy_thread/do_execve/search_binary_handler/->load_binary/do_coredump.

   - sys_fork()/sys_vfork()/sys_clone() unified; some architectures
     still need wrappers (ones with callee-saved registers not saved in
     pt_regs on syscall entry), but the main part of those suckers is in
     kernel/fork.c now."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/signal: (113 commits)
  do_coredump(): get rid of pt_regs argument
  print_fatal_signal(): get rid of pt_regs argument
  ptrace_signal(): get rid of unused arguments
  get rid of ptrace_signal_deliver() arguments
  new helper: signal_pt_regs()
  unify default ptrace_signal_deliver
  flagday: kill pt_regs argument of do_fork()
  death to idle_regs()
  don't pass regs to copy_process()
  flagday: don't pass regs to copy_thread()
  bfin: switch to generic vfork, get rid of pointless wrappers
  xtensa: switch to generic clone()
  openrisc: switch to use of generic fork and clone
  unicore32: switch to generic clone(2)
  score: switch to generic fork/vfork/clone
  c6x: sanitize copy_thread(), get rid of clone(2) wrapper, switch to generic clone()
  take sys_fork/sys_vfork/sys_clone prototypes to linux/syscalls.h
  mn10300: switch to generic fork/vfork/clone
  h8300: switch to generic fork/vfork/clone
  tile: switch to generic clone()
  ...

Conflicts:
	arch/microblaze/include/asm/Kbuild

9 files changed, 67 insertions, 227 deletions

diff --git a/arch/cris/Kconfig b/arch/cris/Kconfig
index a672444..0cac6a4 100644
--- a/arch/cris/Kconfig
+++ b/arch/cris/Kconfig
@@ -49,6 +49,9 @@ config CRIS
 	select GENERIC_SMP_IDLE_THREAD if ETRAX_ARCH_V32
 	select GENERIC_CMOS_UPDATE
 	select MODULES_USE_ELF_RELA
+	select GENERIC_KERNEL_THREAD
+	select GENERIC_KERNEL_EXECVE
+	select CLONE_BACKWARDS2
 
 config HZ
 	int
diff --git a/arch/cris/arch-v10/kernel/entry.S b/arch/cris/arch-v10/kernel/entry.S
index 592fbe9..897bba6 100644
--- a/arch/cris/arch-v10/kernel/entry.S
+++ b/arch/cris/arch-v10/kernel/entry.S
@@ -35,6 +35,7 @@
 	.globl system_call
 	.globl ret_from_intr
 	.globl ret_from_fork
+	.globl ret_from_kernel_thread
 	.globl resume
 	.globl multiple_interrupt
 	.globl hwbreakpoint
@@ -81,7 +82,14 @@ ret_from_fork:
 	jsr schedule_tail
 	ba  ret_from_sys_call
 	nop
-		
+
+ret_from_kernel_thread:
+	jsr schedule_tail
+	move.d	$r2, $r10	; argument is here
+	jsr	$r1		; call the payload
+	moveq	0, $r9		; no syscall restarts, TYVM...
+	ba  ret_from_sys_call
+
 ret_from_intr:
 	;; check for resched if preemptive kernel or if we're going back to user-mode 
 	;; this test matches the user_regs(regs) macro
@@ -586,13 +594,6 @@ _ugdb_handle_breakpoint:
 	ba	do_sigtrap		; SIGTRAP the offending process. 
 	pop	$dccr			; Restore dccr in delay slot.
 
-	.global kernel_execve
-kernel_execve:
-	move.d __NR_execve, $r9
-	break 13
-	ret
-	nop
-
 	.data
 
 hw_bp_trigs:
diff --git a/arch/cris/arch-v10/kernel/process.c b/arch/cris/arch-v10/kernel/process.c
index 15ac715..b101875 100644
--- a/arch/cris/arch-v10/kernel/process.c
+++ b/arch/cris/arch-v10/kernel/process.c
@@ -17,6 +17,7 @@
 #include <arch/svinto.h>
 #include <linux/init.h>
 #include <arch/system.h>
+#include <linux/ptrace.h>
 
 #ifdef CONFIG_ETRAX_GPIO
 void etrax_gpio_wake_up_check(void); /* drivers/gpio.c */
@@ -81,31 +82,6 @@ unsigned long thread_saved_pc(struct task_struct *t)
 	return task_pt_regs(t)->irp;
 }
 
-static void kernel_thread_helper(void* dummy, int (*fn)(void *), void * arg)
-{
-  fn(arg);
-  do_exit(-1); /* Should never be called, return bad exit value */
-}
-
-/*
- * Create a kernel thread
- */
-int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
-{
-	struct pt_regs regs;
-
-	memset(&regs, 0, sizeof(regs));
-
-        /* Don't use r10 since that is set to 0 in copy_thread */
-	regs.r11 = (unsigned long)fn;
-	regs.r12 = (unsigned long)arg;
-	regs.irp = (unsigned long)kernel_thread_helper;
-	regs.dccr = 1 << I_DCCR_BITNR;
-
-	/* Ok, create the new process.. */
-        return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, &regs, 0, NULL, NULL);
-}
-
 /* setup the child's kernel stack with a pt_regs and switch_stack on it.
  * it will be un-nested during _resume and _ret_from_sys_call when the
  * new thread is scheduled.
@@ -115,29 +91,34 @@ int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
  *
  */
 asmlinkage void ret_from_fork(void);
+asmlinkage void ret_from_kernel_thread(void);
 
 int copy_thread(unsigned long clone_flags, unsigned long usp,
-		unsigned long unused,
-		struct task_struct *p, struct pt_regs *regs)
+		unsigned long arg, struct task_struct *p)
 {
-	struct pt_regs * childregs;
-	struct switch_stack *swstack;
+	struct pt_regs *childregs = task_pt_regs(p);
+	struct switch_stack *swstack = ((struct switch_stack *)childregs) - 1;
 	
 	/* put the pt_regs structure at the end of the new kernel stack page and fix it up
 	 * remember that the task_struct doubles as the kernel stack for the task
 	 */
 
-	childregs = task_pt_regs(p);
-        
-	*childregs = *regs;  /* struct copy of pt_regs */
-        
-        p->set_child_tid = p->clear_child_tid = NULL;
+	if (unlikely(p->flags & PF_KTHREAD)) {
+		memset(swstack, 0,
+			sizeof(struct switch_stack) + sizeof(struct pt_regs));
+		swstack->r1 = usp;
+		swstack->r2 = arg;
+		childregs->dccr = 1 << I_DCCR_BITNR;
+		swstack->return_ip = (unsigned long) ret_from_kernel_thread;
+		p->thread.ksp = (unsigned long) swstack;
+		p->thread.usp = 0;
+		return 0;
+	}
+	*childregs = *current_pt_regs();  /* struct copy of pt_regs */
 
         childregs->r10 = 0;  /* child returns 0 after a fork/clone */
-	
-	/* put the switch stack right below the pt_regs */
 
-	swstack = ((struct switch_stack *)childregs) - 1;
+	/* put the switch stack right below the pt_regs */
 
 	swstack->r9 = 0; /* parameter to ret_from_sys_call, 0 == dont restart the syscall */
 
@@ -147,7 +128,7 @@ int copy_thread(unsigned long clone_flags, unsigned long usp,
 	
 	/* fix the user-mode stackpointer */
 
-	p->thread.usp = usp;	
+	p->thread.usp = usp ?: rdusp();
 
 	/* and the kernel-mode one */
 
@@ -161,70 +142,6 @@ int copy_thread(unsigned long clone_flags, unsigned long usp,
 	return 0;
 }
 
-/* 
- * Be aware of the "magic" 7th argument in the four system-calls below.
- * They need the latest stackframe, which is put as the 7th argument by
- * entry.S. The previous arguments are dummies or actually used, but need
- * to be defined to reach the 7th argument.
- *
- * N.B.: Another method to get the stackframe is to use current_regs(). But
- * it returns the latest stack-frame stacked when going from _user mode_ and
- * some of these (at least sys_clone) are called from kernel-mode sometimes
- * (for example during kernel_thread, above) and thus cannot use it. Thus,
- * to be sure not to get any surprises, we use the method for the other calls
- * as well.
- */
-
-asmlinkage int sys_fork(long r10, long r11, long r12, long r13, long mof, long srp,
-			struct pt_regs *regs)
-{
-	return do_fork(SIGCHLD, rdusp(), regs, 0, NULL, NULL);
-}
-
-/* if newusp is 0, we just grab the old usp */
-/* FIXME: Is parent_tid/child_tid really third/fourth argument? Update lib? */
-asmlinkage int sys_clone(unsigned long newusp, unsigned long flags,
-			 int* parent_tid, int* child_tid, long mof, long srp,
-			 struct pt_regs *regs)
-{
-	if (!newusp)
-		newusp = rdusp();
-	return do_fork(flags, newusp, regs, 0, parent_tid, child_tid);
-}
-
-/* vfork is a system call in i386 because of register-pressure - maybe
- * we can remove it and handle it in libc but we put it here until then.
- */
-
-asmlinkage int sys_vfork(long r10, long r11, long r12, long r13, long mof, long srp,
-			 struct pt_regs *regs)
-{
-        return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, rdusp(), regs, 0, NULL, NULL);
-}
-
-/*
- * sys_execve() executes a new program.
- */
-asmlinkage int sys_execve(const char *fname,
-			  const char *const *argv,
-			  const char *const *envp,
-			  long r13, long mof, long srp, 
-			  struct pt_regs *regs)
-{
-	int error;
-	struct filename *filename;
-
-	filename = getname(fname);
-	error = PTR_ERR(filename);
-
-	if (IS_ERR(filename))
-	        goto out;
-	error = do_execve(filename->name, argv, envp, regs);
-	putname(filename);
- out:
-	return error;
-}
-
 unsigned long get_wchan(struct task_struct *p)
 {
 #if 0
diff --git a/arch/cris/arch-v32/kernel/entry.S b/arch/cris/arch-v32/kernel/entry.S
index c3ea4694..faa6441 100644
--- a/arch/cris/arch-v32/kernel/entry.S
+++ b/arch/cris/arch-v32/kernel/entry.S
@@ -31,6 +31,7 @@
 	.globl system_call
 	.globl ret_from_intr
 	.globl ret_from_fork
+	.globl ret_from_kernel_thread
 	.globl resume
 	.globl multiple_interrupt
 	.globl nmi_interrupt
@@ -84,6 +85,18 @@ ret_from_fork:
 	nop
 	.size	ret_from_fork, . - ret_from_fork
 
+	.type	ret_from_kernel_thread,@function
+ret_from_kernel_thread:
+	jsr schedule_tail
+	nop
+	move.d	$r2, $r10
+	jsr	$r1
+	nop
+	moveq	0, $r9			; no syscall restarts, TYVM...
+	ba  ret_from_sys_call
+	nop
+	.size	ret_from_kernel_thread, . - ret_from_kernel_thread
+
 	.type	ret_from_intr,@function
 ret_from_intr:
 	;; Check for resched if preemptive kernel, or if we're going back to
@@ -531,15 +544,6 @@ _ugdb_handle_exception:
 	ba	do_sigtrap		; SIGTRAP the offending process.
 	move.d	[$sp+], $r0		; Restore R0 in delay slot.
 
-	.global kernel_execve
-	.type	kernel_execve,@function
-kernel_execve:
-	move.d __NR_execve, $r9
-	break 13
-	ret
-	nop
-	.size	kernel_execve, . - kernel_execve
-
 	.data
 
 	.section .rodata,"a"
diff --git a/arch/cris/arch-v32/kernel/process.c b/arch/cris/arch-v32/kernel/process.c
index 4e99922..2b23ef0 100644
--- a/arch/cris/arch-v32/kernel/process.c
+++ b/arch/cris/arch-v32/kernel/process.c
@@ -16,6 +16,7 @@
 #include <hwregs/reg_map.h>
 #include <hwregs/timer_defs.h>
 #include <hwregs/intr_vect_defs.h>
+#include <linux/ptrace.h>
 
 extern void stop_watchdog(void);
 
@@ -94,31 +95,6 @@ unsigned long thread_saved_pc(struct task_struct *t)
 	return task_pt_regs(t)->erp;
 }
 
-static void
-kernel_thread_helper(void* dummy, int (*fn)(void *), void * arg)
-{
-	fn(arg);
-	do_exit(-1); /* Should never be called, return bad exit value. */
-}
-
-/* Create a kernel thread. */
-int
-kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
-{
-	struct pt_regs regs;
-
-	memset(&regs, 0, sizeof(regs));
-
-        /* Don't use r10 since that is set to 0 in copy_thread. */
-	regs.r11 = (unsigned long) fn;
-	regs.r12 = (unsigned long) arg;
-	regs.erp = (unsigned long) kernel_thread_helper;
-	regs.ccs = 1 << (I_CCS_BITNR + CCS_SHIFT);
-
-	/* Create the new process. */
-        return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, &regs, 0, NULL, NULL);
-}
-
 /*
  * Setup the child's kernel stack with a pt_regs and call switch_stack() on it.
  * It will be unnested during _resume and _ret_from_sys_call when the new thread
@@ -129,34 +105,42 @@ kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
  */
 
 extern asmlinkage void ret_from_fork(void);
+extern asmlinkage void ret_from_kernel_thread(void);
 
 int
 copy_thread(unsigned long clone_flags, unsigned long usp,
-	unsigned long unused,
-	struct task_struct *p, struct pt_regs *regs)
+	unsigned long arg, struct task_struct *p)
 {
-	struct pt_regs *childregs;
-	struct switch_stack *swstack;
+	struct pt_regs *childregs = task_pt_regs(p);
+	struct switch_stack *swstack = ((struct switch_stack *) childregs) - 1;
 
 	/*
 	 * Put the pt_regs structure at the end of the new kernel stack page and
 	 * fix it up. Note: the task_struct doubles as the kernel stack for the
 	 * task.
 	 */
-	childregs = task_pt_regs(p);
-	*childregs = *regs;	/* Struct copy of pt_regs. */
-        p->set_child_tid = p->clear_child_tid = NULL;
+	if (unlikely(p->flags & PF_KTHREAD)) {
+		memset(swstack, 0,
+			sizeof(struct switch_stack) + sizeof(struct pt_regs));
+		swstack->r1 = usp;
+		swstack->r2 = arg;
+		childregs->ccs = 1 << (I_CCS_BITNR + CCS_SHIFT);
+		swstack->return_ip = (unsigned long) ret_from_kernel_thread;
+		p->thread.ksp = (unsigned long) swstack;
+		p->thread.usp = 0;
+		return 0;
+	}
+	*childregs = *current_pt_regs();	/* Struct copy of pt_regs. */
         childregs->r10 = 0;	/* Child returns 0 after a fork/clone. */
 
 	/* Set a new TLS ?
 	 * The TLS is in $mof because it is the 5th argument to sys_clone.
 	 */
 	if (p->mm && (clone_flags & CLONE_SETTLS)) {
-		task_thread_info(p)->tls = regs->mof;
+		task_thread_info(p)->tls = childregs->mof;
 	}
 
 	/* Put the switch stack right below the pt_regs. */
-	swstack = ((struct switch_stack *) childregs) - 1;
 
 	/* Parameter to ret_from_sys_call. 0 is don't restart the syscall. */
 	swstack->r9 = 0;
@@ -168,76 +152,12 @@ copy_thread(unsigned long clone_flags, unsigned long usp,
 	swstack->return_ip = (unsigned long) ret_from_fork;
 
 	/* Fix the user-mode and kernel-mode stackpointer. */
-	p->thread.usp = usp;
+	p->thread.usp = usp ?: rdusp();
 	p->thread.ksp = (unsigned long) swstack;
 
 	return 0;
 }
 
-/*
- * Be aware of the "magic" 7th argument in the four system-calls below.
- * They need the latest stackframe, which is put as the 7th argument by
- * entry.S. The previous arguments are dummies or actually used, but need
- * to be defined to reach the 7th argument.
- *
- * N.B.: Another method to get the stackframe is to use current_regs(). But
- * it returns the latest stack-frame stacked when going from _user mode_ and
- * some of these (at least sys_clone) are called from kernel-mode sometimes
- * (for example during kernel_thread, above) and thus cannot use it. Thus,
- * to be sure not to get any surprises, we use the method for the other calls
- * as well.
- */
-asmlinkage int
-sys_fork(long r10, long r11, long r12, long r13, long mof, long srp,
-	struct pt_regs *regs)
-{
-	return do_fork(SIGCHLD, rdusp(), regs, 0, NULL, NULL);
-}
-
-/* FIXME: Is parent_tid/child_tid really third/fourth argument? Update lib? */
-asmlinkage int
-sys_clone(unsigned long newusp, unsigned long flags, int *parent_tid, int *child_tid,
-	unsigned long tls, long srp, struct pt_regs *regs)
-{
-	if (!newusp)
-		newusp = rdusp();
-
-	return do_fork(flags, newusp, regs, 0, parent_tid, child_tid);
-}
-
-/*
- * vfork is a system call in i386 because of register-pressure - maybe
- * we can remove it and handle it in libc but we put it here until then.
- */
-asmlinkage int
-sys_vfork(long r10, long r11, long r12, long r13, long mof, long srp,
-	struct pt_regs *regs)
-{
-	return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, rdusp(), regs, 0, NULL, NULL);
-}
-
-/* sys_execve() executes a new program. */
-asmlinkage int
-sys_execve(const char *fname,
-	   const char *const *argv,
-	   const char *const *envp, long r13, long mof, long srp,
-	   struct pt_regs *regs)
-{
-	int error;
-	struct filename *filename;
-
-	filename = getname(fname);
-	error = PTR_ERR(filename);
-
-	if (IS_ERR(filename))
-	        goto out;
-
-	error = do_execve(filename->name, argv, envp, regs);
-	putname(filename);
- out:
-	return error;
-}
-
 unsigned long
 get_wchan(struct task_struct *p)
 {
diff --git a/arch/cris/include/asm/processor.h b/arch/cris/include/asm/processor.h
index ef4e1bc..675823f 100644
--- a/arch/cris/include/asm/processor.h
+++ b/arch/cris/include/asm/processor.h
@@ -49,8 +49,6 @@ struct task_struct;
 #define task_pt_regs(task) user_regs(task_thread_info(task))
 #define current_regs() task_pt_regs(current)
 
-extern int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags);
-
 unsigned long get_wchan(struct task_struct *p);
 
 #define KSTK_ESP(tsk)   ((tsk) == current ? rdusp() : (tsk)->thread.usp)
diff --git a/arch/cris/include/asm/signal.h b/arch/cris/include/asm/signal.h
index ea6af9a..72dbbf5 100644
--- a/arch/cris/include/asm/signal.h
+++ b/arch/cris/include/asm/signal.h
@@ -152,12 +152,6 @@ typedef struct sigaltstack {
 
 #ifdef __KERNEL__
 #include <asm/sigcontext.h>
-
-/* here we could define asm-optimized sigaddset, sigdelset etc. operations. 
- * if we don't, generic ones are used from linux/signal.h
- */
-#define ptrace_signal_deliver(regs, cookie) do { } while (0)
-
 #endif /* __KERNEL__ */
 
 #endif
diff --git a/arch/cris/include/asm/unistd.h b/arch/cris/include/asm/unistd.h
index 51873a4..f27b542 100644
--- a/arch/cris/include/asm/unistd.h
+++ b/arch/cris/include/asm/unistd.h
@@ -371,6 +371,10 @@
 #define __ARCH_WANT_SYS_SIGPROCMASK
 #define __ARCH_WANT_SYS_RT_SIGACTION
 #define __ARCH_WANT_SYS_RT_SIGSUSPEND
+#define __ARCH_WANT_SYS_EXECVE
+#define __ARCH_WANT_SYS_FORK
+#define __ARCH_WANT_SYS_VFORK
+#define __ARCH_WANT_SYS_CLONE
 
 /*
  * "Conditional" syscalls
diff --git a/arch/cris/kernel/crisksyms.c b/arch/cris/kernel/crisksyms.c
index 7ac000f..5868cee 100644
--- a/arch/cris/kernel/crisksyms.c
+++ b/arch/cris/kernel/crisksyms.c
@@ -30,7 +30,6 @@ extern void __negdi2(void);
 extern void iounmap(volatile void * __iomem);
 
 /* Platform dependent support */
-EXPORT_SYMBOL(kernel_thread);
 EXPORT_SYMBOL(get_cmos_time);
 EXPORT_SYMBOL(loops_per_usec);