diff --git a/arch/x86/include/asm/calling.h b/arch/x86/include/asm/calling.h
index 4b5f7bf2b780..1c8b50edb2db 100644
--- a/arch/x86/include/asm/calling.h
+++ b/arch/x86/include/asm/calling.h
@@ -151,6 +151,15 @@ For 32-bit we have the following conventions - kernel is built with
 	movq_cfi_restore 5*8+\offset, rbx
 	.endm
 
+	.macro ZERO_EXTRA_REGS
+	xorl	%r15d, %r15d
+	xorl	%r14d, %r14d
+	xorl	%r13d, %r13d
+	xorl	%r12d, %r12d
+	xorl	%ebp, %ebp
+	xorl	%ebx, %ebx
+	.endm
+
 	.macro RESTORE_C_REGS_HELPER rstor_rax=1, rstor_rcx=1, rstor_r11=1, rstor_r8910=1, rstor_rdx=1
 	.if \rstor_r11
 	movq_cfi_restore 6*8, r11
diff --git a/arch/x86/include/asm/elf.h b/arch/x86/include/asm/elf.h
index ca3347a9dab5..3563107b5060 100644
--- a/arch/x86/include/asm/elf.h
+++ b/arch/x86/include/asm/elf.h
@@ -171,10 +171,11 @@ do {						\
 static inline void elf_common_init(struct thread_struct *t,
 				   struct pt_regs *regs, const u16 ds)
 {
-	regs->ax = regs->bx = regs->cx = regs->dx = 0;
-	regs->si = regs->di = regs->bp = 0;
+	/* Commented-out registers are cleared in stub_execve */
+	/*regs->ax = regs->bx =*/ regs->cx = regs->dx = 0;
+	regs->si = regs->di /*= regs->bp*/ = 0;
 	regs->r8 = regs->r9 = regs->r10 = regs->r11 = 0;
-	regs->r12 = regs->r13 = regs->r14 = regs->r15 = 0;
+	/*regs->r12 = regs->r13 = regs->r14 = regs->r15 = 0;*/
 	t->fs = t->gs = 0;
 	t->fsindex = t->gsindex = 0;
 	t->ds = t->es = ds;
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 386375d43d14..f4270ff73f2a 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -419,25 +419,27 @@ END(stub_\func)
 
 ENTRY(stub_execve)
 	CFI_STARTPROC
-	addq $8, %rsp
-	DEFAULT_FRAME 0
-	SAVE_EXTRA_REGS
-	call sys_execve
-	movq %rax,RAX(%rsp)
-	RESTORE_EXTRA_REGS
-	jmp int_ret_from_sys_call
+	DEFAULT_FRAME 0, 8
+	call	sys_execve
+return_from_execve:
+	testl	%eax, %eax
+	jz	1f
+	/* exec failed, can use fast SYSRET code path in this case */
+	ret
+1:
+	/* must use IRET code path (pt_regs->cs may have changed) */
+	addq	$8, %rsp
+	ZERO_EXTRA_REGS
+	movq	%rax,RAX(%rsp)
+	jmp	int_ret_from_sys_call
 	CFI_ENDPROC
 END(stub_execve)
 
 ENTRY(stub_execveat)
 	CFI_STARTPROC
-	addq $8, %rsp
-	DEFAULT_FRAME 0
-	SAVE_EXTRA_REGS
-	call sys_execveat
-	movq %rax,RAX(%rsp)
-	RESTORE_EXTRA_REGS
-	jmp int_ret_from_sys_call
+	DEFAULT_FRAME 0, 8
+	call	sys_execveat
+	jmp	return_from_execve
 	CFI_ENDPROC
 END(stub_execveat)
 
@@ -472,25 +474,17 @@ END(stub_x32_rt_sigreturn)
 
 ENTRY(stub_x32_execve)
 	CFI_STARTPROC
-	addq $8, %rsp
-	DEFAULT_FRAME 0
-	SAVE_EXTRA_REGS
-	call compat_sys_execve
-	movq %rax,RAX(%rsp)
-	RESTORE_EXTRA_REGS
-	jmp int_ret_from_sys_call
+	DEFAULT_FRAME 0, 8
+	call	compat_sys_execve
+	jmp	return_from_execve
 	CFI_ENDPROC
 END(stub_x32_execve)
 
 ENTRY(stub_x32_execveat)
 	CFI_STARTPROC
-	addq $8, %rsp
-	DEFAULT_FRAME 0
-	SAVE_EXTRA_REGS
-	call compat_sys_execveat
-	movq %rax,RAX(%rsp)
-	RESTORE_EXTRA_REGS
-	jmp int_ret_from_sys_call
+	DEFAULT_FRAME 0, 8
+	call	compat_sys_execveat
+	jmp	return_from_execve
 	CFI_ENDPROC
 END(stub_x32_execveat)