20 files changed, 47 insertions, 123 deletions
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 037c4e30c27..9195fd80e11 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -112,6 +112,7 @@ config X86
 	select GENERIC_KERNEL_EXECVE
 	select MODULES_USE_ELF_REL if X86_32
 	select MODULES_USE_ELF_RELA if X86_64
+	select CLONE_BACKWARDS if X86_32
 
 config INSTRUCTION_DECODER
 	def_bool y
diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c
index 07b3a68d2d2..a703af19c28 100644
--- a/arch/x86/ia32/ia32_aout.c
+++ b/arch/x86/ia32/ia32_aout.c
@@ -35,7 +35,7 @@
 #undef WARN_OLD
 #undef CORE_DUMP /* definitely broken */
 
-static int load_aout_binary(struct linux_binprm *, struct pt_regs *regs);
+static int load_aout_binary(struct linux_binprm *);
 static int load_aout_library(struct file *);
 
 #ifdef CORE_DUMP
@@ -260,9 +260,10 @@ static u32 __user *create_aout_tables(char __user *p, struct linux_binprm *bprm)
  * These are the functions used to load a.out style executables and shared
  * libraries.  There is no binary dependent code anywhere else.
  */
-static int load_aout_binary(struct linux_binprm *bprm, struct pt_regs *regs)
+static int load_aout_binary(struct linux_binprm *bprm)
 {
 	unsigned long error, fd_offset, rlim;
+	struct pt_regs *regs = current_pt_regs();
 	struct exec ex;
 	int retval;
 
diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S
index 076745fc804..32e6f05ddaa 100644
--- a/arch/x86/ia32/ia32entry.S
+++ b/arch/x86/ia32/ia32entry.S
@@ -467,11 +467,16 @@ GLOBAL(\label)
 	PTREGSCALL stub32_sigaltstack, sys32_sigaltstack, %rdx
 	PTREGSCALL stub32_execve, compat_sys_execve, %rcx
 	PTREGSCALL stub32_fork, sys_fork, %rdi
-	PTREGSCALL stub32_clone, sys32_clone, %rdx
 	PTREGSCALL stub32_vfork, sys_vfork, %rdi
 	PTREGSCALL stub32_iopl, sys_iopl, %rsi
 
 	ALIGN
+GLOBAL(stub32_clone)
+	leaq sys_clone(%rip),%rax
+	mov	%r8, %rcx
+	jmp  ia32_ptregs_common	
+
+	ALIGN
 ia32_ptregs_common:
 	popq %r11
 	CFI_ENDPROC
diff --git a/arch/x86/ia32/sys_ia32.c b/arch/x86/ia32/sys_ia32.c
index 86d68d1c880..d0b689ba7be 100644
--- a/arch/x86/ia32/sys_ia32.c
+++ b/arch/x86/ia32/sys_ia32.c
@@ -385,17 +385,6 @@ asmlinkage long sys32_sendfile(int out_fd, int in_fd,
 	return ret;
 }
 
-asmlinkage long sys32_clone(unsigned int clone_flags, unsigned int newsp,
-			    struct pt_regs *regs)
-{
-	void __user *parent_tid = (void __user *)regs->dx;
-	void __user *child_tid = (void __user *)regs->di;
-
-	if (!newsp)
-		newsp = regs->sp;
-	return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid);
-}
-
 /*
  * Some system calls that need sign extended arguments. This could be
  * done by a generic wrapper.
diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index e101b38912d..888184b2fc8 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -178,8 +178,6 @@ static inline int hlt_works(int cpu)
 
 extern void cpu_detect(struct cpuinfo_x86 *c);
 
-extern struct pt_regs *idle_regs(struct pt_regs *);
-
 extern void early_cpu_init(void);
 extern void identify_boot_cpu(void);
 extern void identify_secondary_cpu(struct cpuinfo_x86 *);
diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h
index 323973f4abf..0dba8b7a6ac 100644
--- a/arch/x86/include/asm/signal.h
+++ b/arch/x86/include/asm/signal.h
@@ -260,8 +260,6 @@ struct pt_regs;
 
 #endif /* !__i386__ */
 
-#define ptrace_signal_deliver(regs, cookie) do { } while (0)
-
 #endif /* __KERNEL__ */
 #endif /* __ASSEMBLY__ */
 
diff --git a/arch/x86/include/asm/sys_ia32.h b/arch/x86/include/asm/sys_ia32.h
index a9a8cf3da49..c76fae4d90b 100644
--- a/arch/x86/include/asm/sys_ia32.h
+++ b/arch/x86/include/asm/sys_ia32.h
@@ -54,8 +54,6 @@ asmlinkage long sys32_pwrite(unsigned int, const char __user *, u32, u32, u32);
 asmlinkage long sys32_personality(unsigned long);
 asmlinkage long sys32_sendfile(int, int, compat_off_t __user *, s32);
 
-asmlinkage long sys32_clone(unsigned int, unsigned int, struct pt_regs *);
-
 long sys32_lseek(unsigned int, int, unsigned int);
 long sys32_kill(int, int);
 long sys32_fadvise64_64(int, __u32, __u32, __u32, __u32, int);
diff --git a/arch/x86/include/asm/syscalls.h b/arch/x86/include/asm/syscalls.h
index 2be0b880417..2f8374718aa 100644
--- a/arch/x86/include/asm/syscalls.h
+++ b/arch/x86/include/asm/syscalls.h
@@ -20,15 +20,6 @@
 asmlinkage long sys_ioperm(unsigned long, unsigned long, int);
 long sys_iopl(unsigned int, struct pt_regs *);
 
-/* kernel/process.c */
-int sys_fork(struct pt_regs *);
-int sys_vfork(struct pt_regs *);
-long sys_execve(const char __user *,
-		const char __user *const __user *,
-		const char __user *const __user *);
-long sys_clone(unsigned long, unsigned long, void __user *,
-	       void __user *, struct pt_regs *);
-
 /* kernel/ldt.c */
 asmlinkage int sys_modify_ldt(int, void __user *, unsigned long);
 
diff --git a/arch/x86/include/asm/unistd.h b/arch/x86/include/asm/unistd.h
index 16f3fc6ebf2..0e7dea7d366 100644
--- a/arch/x86/include/asm/unistd.h
+++ b/arch/x86/include/asm/unistd.h
@@ -51,6 +51,9 @@
 # define __ARCH_WANT_SYS_UTIME
 # define __ARCH_WANT_SYS_WAITPID
 # 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/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index ca165ac6793..9c3ab43a695 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -1173,15 +1173,6 @@ DEFINE_PER_CPU(struct task_struct *, fpu_owner_task);
 DEFINE_PER_CPU_ALIGNED(struct stack_canary, stack_canary);
 #endif
 
-/* Make sure %fs and %gs are initialized properly in idle threads */
-struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs)
-{
-	memset(regs, 0, sizeof(struct pt_regs));
-	regs->fs = __KERNEL_PERCPU;
-	regs->gs = __KERNEL_STACK_CANARY;
-
-	return regs;
-}
 #endif	/* CONFIG_X86_64 */
 
 /*
diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index 88b725aa1d5..c763116c535 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -739,30 +739,12 @@ ENTRY(ptregs_##name) ; \
 ENDPROC(ptregs_##name)
 
 PTREGSCALL1(iopl)
-PTREGSCALL0(fork)
-PTREGSCALL0(vfork)
 PTREGSCALL2(sigaltstack)
 PTREGSCALL0(sigreturn)
 PTREGSCALL0(rt_sigreturn)
 PTREGSCALL2(vm86)
 PTREGSCALL1(vm86old)
 
-/* Clone is an oddball.  The 4th arg is in %edi */
-ENTRY(ptregs_clone)
-	CFI_STARTPROC
-	leal 4(%esp),%eax
-	pushl_cfi %eax
-	pushl_cfi PT_EDI(%eax)
-	movl PT_EDX(%eax),%ecx
-	movl PT_ECX(%eax),%edx
-	movl PT_EBX(%eax),%eax
-	call sys_clone
-	addl $8,%esp
-	CFI_ADJUST_CFA_OFFSET -8
-	ret
-	CFI_ENDPROC
-ENDPROC(ptregs_clone)
-
 .macro FIXUP_ESPFIX_STACK
 /*
  * Switch back for ESPFIX stack to the normal zerobased stack
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 31b46128a63..70641aff0c2 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -845,9 +845,25 @@ ENTRY(\label)
 END(\label)
 	.endm
 
-	PTREGSCALL stub_clone, sys_clone, %r8
-	PTREGSCALL stub_fork, sys_fork, %rdi
-	PTREGSCALL stub_vfork, sys_vfork, %rdi
+	.macro FORK_LIKE func
+ENTRY(stub_\func)
+	CFI_STARTPROC
+	popq	%r11			/* save return address */
+	PARTIAL_FRAME 0
+	SAVE_REST
+	pushq	%r11			/* put it back on stack */
+	FIXUP_TOP_OF_STACK %r11, 8
+	DEFAULT_FRAME 0 8		/* offset 8: return address */
+	call sys_\func
+	RESTORE_TOP_OF_STACK %r11, 8
+	ret $REST_SKIP		/* pop extended registers */
+	CFI_ENDPROC
+END(stub_\func)
+	.endm
+
+	FORK_LIKE  clone
+	FORK_LIKE  fork
+	FORK_LIKE  vfork
 	PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx
 	PTREGSCALL stub_iopl, sys_iopl, %rsi
 
diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 2f99e312187..2ed787f15bf 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -262,36 +262,6 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
 	propagate_user_return_notify(prev_p, next_p);
 }
 
-int sys_fork(struct pt_regs *regs)
-{
-	return do_fork(SIGCHLD, regs->sp, regs, 0, NULL, NULL);
-}
-
-/*
- * This is trivial, and on the face of it looks like it
- * could equally well be done in user mode.
- *
- * Not so, for quite unobvious reasons - register pressure.
- * In user mode vfork() cannot have a stack frame, and if
- * done by calling the "clone()" system call directly, you
- * do not have enough call-clobbered registers to hold all
- * the information you need.
- */
-int sys_vfork(struct pt_regs *regs)
-{
-	return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->sp, regs, 0,
-		       NULL, NULL);
-}
-
-long
-sys_clone(unsigned long clone_flags, unsigned long newsp,
-	  void __user *parent_tid, void __user *child_tid, struct pt_regs *regs)
-{
-	if (!newsp)
-		newsp = regs->sp;
-	return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid);
-}
-
 /*
  * Idle related variables and functions
  */
diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
index 44e0bff38e7..b5a8905785e 100644
--- a/arch/x86/kernel/process_32.c
+++ b/arch/x86/kernel/process_32.c
@@ -128,8 +128,7 @@ void release_thread(struct task_struct *dead_task)
 }
 
 int copy_thread(unsigned long clone_flags, unsigned long sp,
-	unsigned long arg,
-	struct task_struct *p, struct pt_regs *regs)
+	unsigned long arg, struct task_struct *p)
 {
 	struct pt_regs *childregs = task_pt_regs(p);
 	struct task_struct *tsk;
@@ -138,7 +137,7 @@ int copy_thread(unsigned long clone_flags, unsigned long sp,
 	p->thread.sp = (unsigned long) childregs;
 	p->thread.sp0 = (unsigned long) (childregs+1);
 
-	if (unlikely(!regs)) {
+	if (unlikely(p->flags & PF_KTHREAD)) {
 		/* kernel thread */
 		memset(childregs, 0, sizeof(struct pt_regs));
 		p->thread.ip = (unsigned long) ret_from_kernel_thread;
@@ -156,12 +155,13 @@ int copy_thread(unsigned long clone_flags, unsigned long sp,
 		memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps));
 		return 0;
 	}
-	*childregs = *regs;
+	*childregs = *current_pt_regs();
 	childregs->ax = 0;
-	childregs->sp = sp;
+	if (sp)
+		childregs->sp = sp;
 
 	p->thread.ip = (unsigned long) ret_from_fork;
-	task_user_gs(p) = get_user_gs(regs);
+	task_user_gs(p) = get_user_gs(current_pt_regs());
 
 	p->fpu_counter = 0;
 	p->thread.io_bitmap_ptr = NULL;
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 16c6365e2b8..6e68a619496 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -146,8 +146,7 @@ static inline u32 read_32bit_tls(struct task_struct *t, int tls)
 }
 
 int copy_thread(unsigned long clone_flags, unsigned long sp,
-		unsigned long arg,
-	struct task_struct *p, struct pt_regs *regs)
+		unsigned long arg, struct task_struct *p)
 {
 	int err;
 	struct pt_regs *childregs;
@@ -169,7 +168,7 @@ int copy_thread(unsigned long clone_flags, unsigned long sp,
 	savesegment(ds, p->thread.ds);
 	memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps));
 
-	if (unlikely(!regs)) {
+	if (unlikely(p->flags & PF_KTHREAD)) {
 		/* kernel thread */
 		memset(childregs, 0, sizeof(struct pt_regs));
 		childregs->sp = (unsigned long)childregs;
@@ -181,10 +180,11 @@ int copy_thread(unsigned long clone_flags, unsigned long sp,
 		childregs->flags = X86_EFLAGS_IF | X86_EFLAGS_BIT1;
 		return 0;
 	}
-	*childregs = *regs;
+	*childregs = *current_pt_regs();
 
 	childregs->ax = 0;
-	childregs->sp = sp;
+	if (sp)
+		childregs->sp = sp;
 
 	err = -ENOMEM;
 	memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps));
diff --git a/arch/x86/syscalls/syscall_32.tbl b/arch/x86/syscalls/syscall_32.tbl
index a47103fbc69..ee3c220ee50 100644
--- a/arch/x86/syscalls/syscall_32.tbl
+++ b/arch/x86/syscalls/syscall_32.tbl
@@ -8,7 +8,7 @@
 #
 0	i386	restart_syscall		sys_restart_syscall
 1	i386	exit			sys_exit
-2	i386	fork			ptregs_fork			stub32_fork
+2	i386	fork			sys_fork			stub32_fork
 3	i386	read			sys_read
 4	i386	write			sys_write
 5	i386	open			sys_open			compat_sys_open
@@ -126,7 +126,7 @@
 117	i386	ipc			sys_ipc				sys32_ipc
 118	i386	fsync			sys_fsync
 119	i386	sigreturn		ptregs_sigreturn		stub32_sigreturn
-120	i386	clone			ptregs_clone			stub32_clone
+120	i386	clone			sys_clone			stub32_clone
 121	i386	setdomainname		sys_setdomainname
 122	i386	uname			sys_newuname
 123	i386	modify_ldt		sys_modify_ldt
@@ -196,7 +196,7 @@
 187	i386	sendfile		sys_sendfile			sys32_sendfile
 188	i386	getpmsg
 189	i386	putpmsg
-190	i386	vfork			ptregs_vfork			stub32_vfork
+190	i386	vfork			sys_vfork			stub32_vfork
 191	i386	ugetrlimit		sys_getrlimit			compat_sys_getrlimit
 192	i386	mmap2			sys_mmap_pgoff
 193	i386	truncate64		sys_truncate64			sys32_truncate64
diff --git a/arch/x86/um/Kconfig b/arch/x86/um/Kconfig
index b0c30dae9f5..98399704196 100644
--- a/arch/x86/um/Kconfig
+++ b/arch/x86/um/Kconfig
@@ -25,6 +25,7 @@ config X86_32
 	select HAVE_AOUT
 	select ARCH_WANT_IPC_PARSE_VERSION
 	select MODULES_USE_ELF_REL
+	select CLONE_BACKWARDS
 
 config X86_64
 	def_bool 64BIT
diff --git a/arch/x86/um/shared/sysdep/syscalls.h b/arch/x86/um/shared/sysdep/syscalls.h
index ca255a805ed..bd9a89b67e4 100644
--- a/arch/x86/um/shared/sysdep/syscalls.h
+++ b/arch/x86/um/shared/sysdep/syscalls.h
@@ -1,5 +1,3 @@
-extern long sys_clone(unsigned long clone_flags, unsigned long newsp,
-	       void __user *parent_tid, void __user *child_tid);
 #ifdef __i386__
 #include "syscalls_32.h"
 #else
diff --git a/arch/x86/um/sys_call_table_32.c b/arch/x86/um/sys_call_table_32.c
index 232e60504b3..812e98c098e 100644
--- a/arch/x86/um/sys_call_table_32.c
+++ b/arch/x86/um/sys_call_table_32.c
@@ -24,13 +24,10 @@
 
 #define old_mmap sys_old_mmap
 
-#define ptregs_fork sys_fork
 #define ptregs_iopl sys_iopl
 #define ptregs_vm86old sys_vm86old
-#define ptregs_clone i386_clone
 #define ptregs_vm86 sys_vm86
 #define ptregs_sigaltstack sys_sigaltstack
-#define ptregs_vfork sys_vfork
 
 #define __SYSCALL_I386(nr, sym, compat) extern asmlinkage void sym(void) ;
 #include <asm/syscalls_32.h>
diff --git a/arch/x86/um/syscalls_32.c b/arch/x86/um/syscalls_32.c
index db444c7218f..e8bcea99acd 100644
--- a/arch/x86/um/syscalls_32.c
+++ b/arch/x86/um/syscalls_32.c
@@ -6,21 +6,6 @@
 #include <linux/syscalls.h>
 #include <sysdep/syscalls.h>
 
-/*
- * The prototype on i386 is:
- *
- *     int clone(int flags, void * child_stack, int * parent_tidptr, struct user_desc * newtls
- *
- * and the "newtls" arg. on i386 is read by copy_thread directly from the
- * register saved on the stack.
- */
-long i386_clone(unsigned long clone_flags, unsigned long newsp,
-		int __user *parent_tid, void *newtls, int __user *child_tid)
-{
-	return sys_clone(clone_flags, newsp, parent_tid, child_tid);
-}
-
-
 long sys_sigaction(int sig, const struct old_sigaction __user *act,
 			 struct old_sigaction __user *oact)
 {