diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index f6946b81f74a..4fa665b058e3 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -240,6 +240,7 @@ config X86
 	select HAVE_ARCH_KCSAN			if X86_64
 	select X86_FEATURE_NAMES		if PROC_FS
 	select PROC_PID_ARCH_STATUS		if PROC_FS
+	select SET_FS
 	imply IMA_SECURE_AND_OR_TRUSTED_BOOT    if EFI
 
 config INSTRUCTION_DECODER
diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c
index a09fc37ead9d..ca8a657edf59 100644
--- a/arch/x86/ia32/ia32_aout.c
+++ b/arch/x86/ia32/ia32_aout.c
@@ -239,6 +239,7 @@ beyond_if:
 	(regs)->ss = __USER32_DS;
 	regs->r8 = regs->r9 = regs->r10 = regs->r11 =
 	regs->r12 = regs->r13 = regs->r14 = regs->r15 = 0;
+	set_fs(USER_DS);
 	return 0;
 }
 
diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index 82a08b585818..b4625108942e 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -482,6 +482,10 @@ extern unsigned int fpu_user_xstate_size;
 
 struct perf_event;
 
+typedef struct {
+	unsigned long		seg;
+} mm_segment_t;
+
 struct thread_struct {
 	/* Cached TLS descriptors: */
 	struct desc_struct	tls_array[GDT_ENTRY_TLS_ENTRIES];
@@ -534,6 +538,8 @@ struct thread_struct {
 	 */
 	unsigned long		iopl_emul;
 
+	mm_segment_t		addr_limit;
+
 	unsigned int		sig_on_uaccess_err:1;
 
 	/* Floating point and extended processor state */
@@ -780,12 +786,15 @@ static inline void spin_lock_prefetch(const void *x)
 #define INIT_THREAD  {							  \
 	.sp0			= TOP_OF_INIT_STACK,			  \
 	.sysenter_cs		= __KERNEL_CS,				  \
+	.addr_limit		= KERNEL_DS,				  \
 }
 
 #define KSTK_ESP(task)		(task_pt_regs(task)->sp)
 
 #else
-#define INIT_THREAD { }
+#define INIT_THREAD  {						\
+	.addr_limit		= KERNEL_DS,			\
+}
 
 extern unsigned long KSTK_ESP(struct task_struct *task);
 
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index 44733a4bfc42..267701ae3d86 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -102,6 +102,7 @@ struct thread_info {
 #define TIF_SYSCALL_TRACEPOINT	28	/* syscall tracepoint instrumentation */
 #define TIF_ADDR32		29	/* 32-bit address space on 64 bits */
 #define TIF_X32			30	/* 32-bit native x86-64 binary */
+#define TIF_FSCHECK		31	/* Check FS is USER_DS on return */
 
 #define _TIF_SYSCALL_TRACE	(1 << TIF_SYSCALL_TRACE)
 #define _TIF_NOTIFY_RESUME	(1 << TIF_NOTIFY_RESUME)
@@ -130,6 +131,7 @@ struct thread_info {
 #define _TIF_SYSCALL_TRACEPOINT	(1 << TIF_SYSCALL_TRACEPOINT)
 #define _TIF_ADDR32		(1 << TIF_ADDR32)
 #define _TIF_X32		(1 << TIF_X32)
+#define _TIF_FSCHECK		(1 << TIF_FSCHECK)
 
 /* flags to check in __switch_to() */
 #define _TIF_WORK_CTXSW_BASE					\
diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h
index f13659523108..477c503f2753 100644
--- a/arch/x86/include/asm/uaccess.h
+++ b/arch/x86/include/asm/uaccess.h
@@ -12,6 +12,30 @@
 #include <asm/smap.h>
 #include <asm/extable.h>
 
+/*
+ * The fs value determines whether argument validity checking should be
+ * performed or not.  If get_fs() == USER_DS, checking is performed, with
+ * get_fs() == KERNEL_DS, checking is bypassed.
+ *
+ * For historical reasons, these macros are grossly misnamed.
+ */
+
+#define MAKE_MM_SEG(s)	((mm_segment_t) { (s) })
+
+#define KERNEL_DS	MAKE_MM_SEG(-1UL)
+#define USER_DS 	MAKE_MM_SEG(TASK_SIZE_MAX)
+
+#define get_fs()	(current->thread.addr_limit)
+static inline void set_fs(mm_segment_t fs)
+{
+	current->thread.addr_limit = fs;
+	/* On user-mode return, check fs is correct */
+	set_thread_flag(TIF_FSCHECK);
+}
+
+#define uaccess_kernel() (get_fs().seg == KERNEL_DS.seg)
+#define user_addr_max() (current->thread.addr_limit.seg)
+
 /*
  * Test whether a block of memory is a valid user space address.
  * Returns 0 if the range is valid, nonzero otherwise.
@@ -69,7 +93,7 @@ static inline bool pagefault_disabled(void);
 #define access_ok(addr, size)					\
 ({									\
 	WARN_ON_IN_IRQ();						\
-	likely(!__range_not_ok(addr, size, TASK_SIZE_MAX));		\
+	likely(!__range_not_ok(addr, size, user_addr_max()));		\
 })
 
 extern int __get_user_1(void);
diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c
index 70b7154f4bdd..3ca07ad552ae 100644
--- a/arch/x86/kernel/asm-offsets.c
+++ b/arch/x86/kernel/asm-offsets.c
@@ -37,6 +37,9 @@ static void __used common(void)
 	OFFSET(TASK_stack_canary, task_struct, stack_canary);
 #endif
 
+	BLANK();
+	OFFSET(TASK_addr_limit, task_struct, thread.addr_limit);
+
 	BLANK();
 	OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx);
 
diff --git a/arch/x86/lib/getuser.S b/arch/x86/lib/getuser.S
index fa1bc2104b32..2cd902e06062 100644
--- a/arch/x86/lib/getuser.S
+++ b/arch/x86/lib/getuser.S
@@ -37,19 +37,10 @@
 
 #define ASM_BARRIER_NOSPEC ALTERNATIVE "", "lfence", X86_FEATURE_LFENCE_RDTSC
 
-#ifdef CONFIG_X86_5LEVEL
-#define LOAD_TASK_SIZE_MINUS_N(n) \
-	ALTERNATIVE __stringify(mov $((1 << 47) - 4096 - (n)),%rdx), \
-		    __stringify(mov $((1 << 56) - 4096 - (n)),%rdx), X86_FEATURE_LA57
-#else
-#define LOAD_TASK_SIZE_MINUS_N(n) \
-	mov $(TASK_SIZE_MAX - (n)),%_ASM_DX
-#endif
-
 	.text
 SYM_FUNC_START(__get_user_1)
-	LOAD_TASK_SIZE_MINUS_N(0)
-	cmp %_ASM_DX,%_ASM_AX
+	mov PER_CPU_VAR(current_task), %_ASM_DX
+	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
 	sbb %_ASM_DX, %_ASM_DX		/* array_index_mask_nospec() */
 	and %_ASM_DX, %_ASM_AX
@@ -62,13 +53,15 @@ SYM_FUNC_END(__get_user_1)
 EXPORT_SYMBOL(__get_user_1)
 
 SYM_FUNC_START(__get_user_2)
-	LOAD_TASK_SIZE_MINUS_N(1)
-	cmp %_ASM_DX,%_ASM_AX
+	add $1,%_ASM_AX
+	jc bad_get_user
+	mov PER_CPU_VAR(current_task), %_ASM_DX
+	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
 	sbb %_ASM_DX, %_ASM_DX		/* array_index_mask_nospec() */
 	and %_ASM_DX, %_ASM_AX
 	ASM_STAC
-2:	movzwl (%_ASM_AX),%edx
+2:	movzwl -1(%_ASM_AX),%edx
 	xor %eax,%eax
 	ASM_CLAC
 	ret
@@ -76,13 +69,15 @@ SYM_FUNC_END(__get_user_2)
 EXPORT_SYMBOL(__get_user_2)
 
 SYM_FUNC_START(__get_user_4)
-	LOAD_TASK_SIZE_MINUS_N(3)
-	cmp %_ASM_DX,%_ASM_AX
+	add $3,%_ASM_AX
+	jc bad_get_user
+	mov PER_CPU_VAR(current_task), %_ASM_DX
+	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
 	sbb %_ASM_DX, %_ASM_DX		/* array_index_mask_nospec() */
 	and %_ASM_DX, %_ASM_AX
 	ASM_STAC
-3:	movl (%_ASM_AX),%edx
+3:	movl -3(%_ASM_AX),%edx
 	xor %eax,%eax
 	ASM_CLAC
 	ret
@@ -91,25 +86,29 @@ EXPORT_SYMBOL(__get_user_4)
 
 SYM_FUNC_START(__get_user_8)
 #ifdef CONFIG_X86_64
-	LOAD_TASK_SIZE_MINUS_N(7)
-	cmp %_ASM_DX,%_ASM_AX
+	add $7,%_ASM_AX
+	jc bad_get_user
+	mov PER_CPU_VAR(current_task), %_ASM_DX
+	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
 	sbb %_ASM_DX, %_ASM_DX		/* array_index_mask_nospec() */
 	and %_ASM_DX, %_ASM_AX
 	ASM_STAC
-4:	movq (%_ASM_AX),%rdx
+4:	movq -7(%_ASM_AX),%rdx
 	xor %eax,%eax
 	ASM_CLAC
 	ret
 #else
-	LOAD_TASK_SIZE_MINUS_N(7)
-	cmp %_ASM_DX,%_ASM_AX
+	add $7,%_ASM_AX
+	jc bad_get_user_8
+	mov PER_CPU_VAR(current_task), %_ASM_DX
+	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user_8
 	sbb %_ASM_DX, %_ASM_DX		/* array_index_mask_nospec() */
 	and %_ASM_DX, %_ASM_AX
 	ASM_STAC
-4:	movl (%_ASM_AX),%edx
-5:	movl 4(%_ASM_AX),%ecx
+4:	movl -7(%_ASM_AX),%edx
+5:	movl -3(%_ASM_AX),%ecx
 	xor %eax,%eax
 	ASM_CLAC
 	ret
diff --git a/arch/x86/lib/putuser.S b/arch/x86/lib/putuser.S
index 0ea344c5ea43..b34a17763f28 100644
--- a/arch/x86/lib/putuser.S
+++ b/arch/x86/lib/putuser.S
@@ -33,19 +33,12 @@
  * as they get called from within inline assembly.
  */
 
-#ifdef CONFIG_X86_5LEVEL
-#define LOAD_TASK_SIZE_MINUS_N(n) \
-	ALTERNATIVE __stringify(mov $((1 << 47) - 4096 - (n)),%rbx), \
-		    __stringify(mov $((1 << 56) - 4096 - (n)),%rbx), X86_FEATURE_LA57
-#else
-#define LOAD_TASK_SIZE_MINUS_N(n) \
-	mov $(TASK_SIZE_MAX - (n)),%_ASM_BX
-#endif
+#define ENTER	mov PER_CPU_VAR(current_task), %_ASM_BX
 
 .text
 SYM_FUNC_START(__put_user_1)
-	LOAD_TASK_SIZE_MINUS_N(0)
-	cmp %_ASM_BX,%_ASM_CX
+	ENTER
+	cmp TASK_addr_limit(%_ASM_BX),%_ASM_CX
 	jae .Lbad_put_user
 SYM_INNER_LABEL(__put_user_nocheck_1, SYM_L_GLOBAL)
 	ASM_STAC
@@ -58,7 +51,9 @@ EXPORT_SYMBOL(__put_user_1)
 EXPORT_SYMBOL(__put_user_nocheck_1)
 
 SYM_FUNC_START(__put_user_2)
-	LOAD_TASK_SIZE_MINUS_N(1)
+	ENTER
+	mov TASK_addr_limit(%_ASM_BX),%_ASM_BX
+	sub $1,%_ASM_BX
 	cmp %_ASM_BX,%_ASM_CX
 	jae .Lbad_put_user
 SYM_INNER_LABEL(__put_user_nocheck_2, SYM_L_GLOBAL)
@@ -72,7 +67,9 @@ EXPORT_SYMBOL(__put_user_2)
 EXPORT_SYMBOL(__put_user_nocheck_2)
 
 SYM_FUNC_START(__put_user_4)
-	LOAD_TASK_SIZE_MINUS_N(3)
+	ENTER
+	mov TASK_addr_limit(%_ASM_BX),%_ASM_BX
+	sub $3,%_ASM_BX
 	cmp %_ASM_BX,%_ASM_CX
 	jae .Lbad_put_user
 SYM_INNER_LABEL(__put_user_nocheck_4, SYM_L_GLOBAL)
@@ -86,7 +83,9 @@ EXPORT_SYMBOL(__put_user_4)
 EXPORT_SYMBOL(__put_user_nocheck_4)
 
 SYM_FUNC_START(__put_user_8)
-	LOAD_TASK_SIZE_MINUS_N(7)
+	ENTER
+	mov TASK_addr_limit(%_ASM_BX),%_ASM_BX
+	sub $7,%_ASM_BX
 	cmp %_ASM_BX,%_ASM_CX
 	jae .Lbad_put_user
 SYM_INNER_LABEL(__put_user_nocheck_8, SYM_L_GLOBAL)