diff --git a/arch/arm64/include/asm/assembler.h b/arch/arm64/include/asm/assembler.h
index fd3e392..5901480 100644
--- a/arch/arm64/include/asm/assembler.h
+++ b/arch/arm64/include/asm/assembler.h
@@ -21,6 +21,7 @@
 #endif
 
 #include <asm/ptrace.h>
+#include <asm/thread_info.h>
 
 /*
  * Stack pushing/popping (register pairs only). Equivalent to store decrement
@@ -68,23 +69,31 @@
 	msr	daifclr, #8
 	.endm
 
-	.macro	disable_step, tmp
+	.macro	disable_step_tsk, flgs, tmp
+	tbz	\flgs, #TIF_SINGLESTEP, 9990f
 	mrs	\tmp, mdscr_el1
 	bic	\tmp, \tmp, #1
 	msr	mdscr_el1, \tmp
+	isb	// Synchronise with enable_dbg
+9990:
 	.endm
 
-	.macro	enable_step, tmp
+	.macro	enable_step_tsk, flgs, tmp
+	tbz	\flgs, #TIF_SINGLESTEP, 9990f
+	disable_dbg
 	mrs	\tmp, mdscr_el1
 	orr	\tmp, \tmp, #1
 	msr	mdscr_el1, \tmp
+9990:
 	.endm
 
-	.macro	enable_dbg_if_not_stepping, tmp
-	mrs	\tmp, mdscr_el1
-	tbnz	\tmp, #0, 9990f
-	enable_dbg
-9990:
+/*
+ * Enable both debug exceptions and interrupts. This is likely to be
+ * faster than two daifclr operations, since writes to this register
+ * are self-synchronising.
+ */
+	.macro	enable_dbg_and_irq
+	msr	daifclr, #(8 | 2)
 	.endm
 
 /*
diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S
index 39ac630..e8b23a3 100644
--- a/arch/arm64/kernel/entry.S
+++ b/arch/arm64/kernel/entry.S
@@ -60,6 +60,9 @@
 	push	x0, x1
 	.if	\el == 0
 	mrs	x21, sp_el0
+	get_thread_info tsk			// Ensure MDSCR_EL1.SS is clear,
+	ldr	x19, [tsk, #TI_FLAGS]		// since we can unmask debug
+	disable_step_tsk x19, x20		// exceptions when scheduling.
 	.else
 	add	x21, sp, #S_FRAME_SIZE
 	.endif
@@ -259,7 +262,7 @@
 	 * Data abort handling
 	 */
 	mrs	x0, far_el1
-	enable_dbg_if_not_stepping x2
+	enable_dbg
 	// re-enable interrupts if they were enabled in the aborted context
 	tbnz	x23, #7, 1f			// PSR_I_BIT
 	enable_irq
@@ -275,6 +278,7 @@
 	 * Stack or PC alignment exception handling
 	 */
 	mrs	x0, far_el1
+	enable_dbg
 	mov	x1, x25
 	mov	x2, sp
 	b	do_sp_pc_abort
@@ -282,6 +286,7 @@
 	/*
 	 * Undefined instruction
 	 */
+	enable_dbg
 	mov	x0, sp
 	b	do_undefinstr
 el1_dbg:
@@ -294,10 +299,11 @@
 	mrs	x0, far_el1
 	mov	x2, sp				// struct pt_regs
 	bl	do_debug_exception
-
+	enable_dbg
 	kernel_exit 1
 el1_inv:
 	// TODO: add support for undefined instructions in kernel mode
+	enable_dbg
 	mov	x0, sp
 	mov	x1, #BAD_SYNC
 	mrs	x2, esr_el1
@@ -307,7 +313,7 @@
 	.align	6
 el1_irq:
 	kernel_entry 1
-	enable_dbg_if_not_stepping x0
+	enable_dbg
 #ifdef CONFIG_TRACE_IRQFLAGS
 	bl	trace_hardirqs_off
 #endif
@@ -332,8 +338,7 @@
 #ifdef CONFIG_PREEMPT
 el1_preempt:
 	mov	x24, lr
-1:	enable_dbg
-	bl	preempt_schedule_irq		// irq en/disable is done inside
+1:	bl	preempt_schedule_irq		// irq en/disable is done inside
 	ldr	x0, [tsk, #TI_FLAGS]		// get new tasks TI_FLAGS
 	tbnz	x0, #TIF_NEED_RESCHED, 1b	// needs rescheduling?
 	ret	x24
@@ -349,7 +354,7 @@
 	lsr	x24, x25, #ESR_EL1_EC_SHIFT	// exception class
 	cmp	x24, #ESR_EL1_EC_SVC64		// SVC in 64-bit state
 	b.eq	el0_svc
-	adr	lr, ret_from_exception
+	adr	lr, ret_to_user
 	cmp	x24, #ESR_EL1_EC_DABT_EL0	// data abort in EL0
 	b.eq	el0_da
 	cmp	x24, #ESR_EL1_EC_IABT_EL0	// instruction abort in EL0
@@ -378,7 +383,7 @@
 	lsr	x24, x25, #ESR_EL1_EC_SHIFT	// exception class
 	cmp	x24, #ESR_EL1_EC_SVC32		// SVC in 32-bit state
 	b.eq	el0_svc_compat
-	adr	lr, ret_from_exception
+	adr	lr, ret_to_user
 	cmp	x24, #ESR_EL1_EC_DABT_EL0	// data abort in EL0
 	b.eq	el0_da
 	cmp	x24, #ESR_EL1_EC_IABT_EL0	// instruction abort in EL0
@@ -423,11 +428,8 @@
 	 */
 	mrs	x0, far_el1
 	bic	x0, x0, #(0xff << 56)
-	disable_step x1
-	isb
-	enable_dbg
 	// enable interrupts before calling the main handler
-	enable_irq
+	enable_dbg_and_irq
 	mov	x1, x25
 	mov	x2, sp
 	b	do_mem_abort
@@ -436,11 +438,8 @@
 	 * Instruction abort handling
 	 */
 	mrs	x0, far_el1
-	disable_step x1
-	isb
-	enable_dbg
 	// enable interrupts before calling the main handler
-	enable_irq
+	enable_dbg_and_irq
 	orr	x1, x25, #1 << 24		// use reserved ISS bit for instruction aborts
 	mov	x2, sp
 	b	do_mem_abort
@@ -448,6 +447,7 @@
 	/*
 	 * Floating Point or Advanced SIMD access
 	 */
+	enable_dbg
 	mov	x0, x25
 	mov	x1, sp
 	b	do_fpsimd_acc
@@ -455,6 +455,7 @@
 	/*
 	 * Floating Point or Advanced SIMD exception
 	 */
+	enable_dbg
 	mov	x0, x25
 	mov	x1, sp
 	b	do_fpsimd_exc
@@ -463,11 +464,8 @@
 	 * Stack or PC alignment exception handling
 	 */
 	mrs	x0, far_el1
-	disable_step x1
-	isb
-	enable_dbg
 	// enable interrupts before calling the main handler
-	enable_irq
+	enable_dbg_and_irq
 	mov	x1, x25
 	mov	x2, sp
 	b	do_sp_pc_abort
@@ -475,9 +473,9 @@
 	/*
 	 * Undefined instruction
 	 */
-	mov	x0, sp
 	// enable interrupts before calling the main handler
-	enable_irq
+	enable_dbg_and_irq
+	mov	x0, sp
 	b	do_undefinstr
 el0_dbg:
 	/*
@@ -485,11 +483,13 @@
 	 */
 	tbnz	x24, #0, el0_inv		// EL0 only
 	mrs	x0, far_el1
-	disable_step x1
 	mov	x1, x25
 	mov	x2, sp
-	b	do_debug_exception
+	bl	do_debug_exception
+	enable_dbg
+	b	ret_to_user
 el0_inv:
+	enable_dbg
 	mov	x0, sp
 	mov	x1, #BAD_SYNC
 	mrs	x2, esr_el1
@@ -500,15 +500,12 @@
 el0_irq:
 	kernel_entry 0
 el0_irq_naked:
-	disable_step x1
-	isb
 	enable_dbg
 #ifdef CONFIG_TRACE_IRQFLAGS
 	bl	trace_hardirqs_off
 #endif
 
 	irq_handler
-	get_thread_info tsk
 
 #ifdef CONFIG_TRACE_IRQFLAGS
 	bl	trace_hardirqs_on
@@ -517,14 +514,6 @@
 ENDPROC(el0_irq)
 
 /*
- * This is the return code to user mode for abort handlers
- */
-ret_from_exception:
-	get_thread_info tsk
-	b	ret_to_user
-ENDPROC(ret_from_exception)
-
-/*
  * Register switch for AArch64. The callee-saved registers need to be saved
  * and restored. On entry:
  *   x0 = previous task_struct (must be preserved across the switch)
@@ -563,10 +552,7 @@
 	ldr	x1, [tsk, #TI_FLAGS]
 	and	x2, x1, #_TIF_WORK_MASK
 	cbnz	x2, fast_work_pending
-	tbz	x1, #TIF_SINGLESTEP, fast_exit
-	disable_dbg
-	enable_step x2
-fast_exit:
+	enable_step_tsk x1, x2
 	kernel_exit 0, ret = 1
 
 /*
@@ -585,7 +571,6 @@
 	bl	do_notify_resume
 	b	ret_to_user
 work_resched:
-	enable_dbg
 	bl	schedule
 
 /*
@@ -596,9 +581,7 @@
 	ldr	x1, [tsk, #TI_FLAGS]
 	and	x2, x1, #_TIF_WORK_MASK
 	cbnz	x2, work_pending
-	tbz	x1, #TIF_SINGLESTEP, no_work_pending
-	disable_dbg
-	enable_step x2
+	enable_step_tsk x1, x2
 no_work_pending:
 	kernel_exit 0, ret = 0
 ENDPROC(ret_to_user)
@@ -625,12 +608,8 @@
 	mov	sc_nr, #__NR_syscalls
 el0_svc_naked:					// compat entry point
 	stp	x0, scno, [sp, #S_ORIG_X0]	// save the original x0 and syscall number
-	disable_step x16
-	isb
-	enable_dbg
-	enable_irq
+	enable_dbg_and_irq
 
-	get_thread_info tsk
 	ldr	x16, [tsk, #TI_FLAGS]		// check for syscall tracing
 	tbnz	x16, #TIF_SYSCALL_TRACE, __sys_trace // are we tracing syscalls?
 	adr	lr, ret_fast_syscall		// return address