Merge branch 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm

Pull ARM updates from Russell King:

 - Initial round of Spectre variant 1 and variant 2 fixes for 32-bit ARM

 - Clang support improvements

 - nommu updates for v8 MPU

 - enable ARM_MODULE_PLTS by default to avoid problems loading modules
   with larger kernels

 - vmlinux.lds and dma-mapping cleanups

* 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm: (31 commits)
  ARM: spectre-v1: fix syscall entry
  ARM: spectre-v1: add array_index_mask_nospec() implementation
  ARM: spectre-v1: add speculation barrier (csdb) macros
  ARM: KVM: report support for SMCCC_ARCH_WORKAROUND_1
  ARM: KVM: Add SMCCC_ARCH_WORKAROUND_1 fast handling
  ARM: spectre-v2: KVM: invalidate icache on guest exit for Brahma B15
  ARM: KVM: invalidate icache on guest exit for Cortex-A15
  ARM: KVM: invalidate BTB on guest exit for Cortex-A12/A17
  ARM: spectre-v2: warn about incorrect context switching functions
  ARM: spectre-v2: add firmware based hardening
  ARM: spectre-v2: harden user aborts in kernel space
  ARM: spectre-v2: add Cortex A8 and A15 validation of the IBE bit
  ARM: spectre-v2: harden branch predictor on context switches
  ARM: spectre: add Kconfig symbol for CPUs vulnerable to Spectre
  ARM: bugs: add support for per-processor bug checking
  ARM: bugs: hook processor bug checking into SMP and suspend paths
  ARM: bugs: prepare processor bug infrastructure
  ARM: add more CPU part numbers for Cortex and Brahma B15 CPUs
  ARM: 8774/1: remove no-op macro VMLINUX_SYMBOL()
  ARM: 8773/1: amba: Export amba_bustype
  ...

11 files changed, 315 insertions, 80 deletions

diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile
index b59ac4bf82b8..8cad59465af3 100644
--- a/arch/arm/kernel/Makefile
+++ b/arch/arm/kernel/Makefile
@@ -31,6 +31,7 @@ else
 obj-y		+= entry-armv.o
 endif
 
+obj-$(CONFIG_MMU)		+= bugs.o
 obj-$(CONFIG_CPU_IDLE)		+= cpuidle.o
 obj-$(CONFIG_ISA_DMA_API)	+= dma.o
 obj-$(CONFIG_FIQ)		+= fiq.o fiqasm.o
diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c
index f369ece99958..27c5381518d8 100644
--- a/arch/arm/kernel/asm-offsets.c
+++ b/arch/arm/kernel/asm-offsets.c
@@ -194,9 +194,11 @@ int main(void)
   DEFINE(MPU_RNG_INFO_USED,	offsetof(struct mpu_rgn_info, used));
 
   DEFINE(MPU_RNG_SIZE,		sizeof(struct mpu_rgn));
-  DEFINE(MPU_RGN_DRBAR,		offsetof(struct mpu_rgn, drbar));
-  DEFINE(MPU_RGN_DRSR,		offsetof(struct mpu_rgn, drsr));
-  DEFINE(MPU_RGN_DRACR,		offsetof(struct mpu_rgn, dracr));
+  DEFINE(MPU_RGN_DRBAR,	offsetof(struct mpu_rgn, drbar));
+  DEFINE(MPU_RGN_DRSR,	offsetof(struct mpu_rgn, drsr));
+  DEFINE(MPU_RGN_DRACR,	offsetof(struct mpu_rgn, dracr));
+  DEFINE(MPU_RGN_PRBAR,	offsetof(struct mpu_rgn, prbar));
+  DEFINE(MPU_RGN_PRLAR,	offsetof(struct mpu_rgn, prlar));
 #endif
   return 0; 
 }
diff --git a/arch/arm/kernel/bugs.c b/arch/arm/kernel/bugs.c
new file mode 100644
index 000000000000..7be511310191
--- /dev/null
+++ b/arch/arm/kernel/bugs.c
@@ -0,0 +1,18 @@
+// SPDX-Identifier: GPL-2.0
+#include <linux/init.h>
+#include <asm/bugs.h>
+#include <asm/proc-fns.h>
+
+void check_other_bugs(void)
+{
+#ifdef MULTI_CPU
+	if (processor.check_bugs)
+		processor.check_bugs();
+#endif
+}
+
+void __init check_bugs(void)
+{
+	check_writebuffer_bugs();
+	check_other_bugs();
+}
diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S
index 3c4f88701f22..20df608bf343 100644
--- a/arch/arm/kernel/entry-common.S
+++ b/arch/arm/kernel/entry-common.S
@@ -242,9 +242,7 @@ local_restart:
 	tst	r10, #_TIF_SYSCALL_WORK		@ are we tracing syscalls?
 	bne	__sys_trace
 
-	cmp	scno, #NR_syscalls		@ check upper syscall limit
-	badr	lr, ret_fast_syscall		@ return address
-	ldrcc	pc, [tbl, scno, lsl #2]		@ call sys_* routine
+	invoke_syscall tbl, scno, r10, ret_fast_syscall
 
 	add	r1, sp, #S_OFF
 2:	cmp	scno, #(__ARM_NR_BASE - __NR_SYSCALL_BASE)
@@ -278,14 +276,8 @@ __sys_trace:
 	mov	r1, scno
 	add	r0, sp, #S_OFF
 	bl	syscall_trace_enter
-
-	badr	lr, __sys_trace_return		@ return address
-	mov	scno, r0			@ syscall number (possibly new)
-	add	r1, sp, #S_R0 + S_OFF		@ pointer to regs
-	cmp	scno, #NR_syscalls		@ check upper syscall limit
-	ldmccia	r1, {r0 - r6}			@ have to reload r0 - r6
-	stmccia	sp, {r4, r5}			@ and update the stack args
-	ldrcc	pc, [tbl, scno, lsl #2]		@ call sys_* routine
+	mov	scno, r0
+	invoke_syscall tbl, scno, r10, __sys_trace_return, reload=1
 	cmp	scno, #-1			@ skip the syscall?
 	bne	2b
 	add	sp, sp, #S_OFF			@ restore stack
@@ -363,6 +355,10 @@ sys_syscall:
 		bic	scno, r0, #__NR_OABI_SYSCALL_BASE
 		cmp	scno, #__NR_syscall - __NR_SYSCALL_BASE
 		cmpne	scno, #NR_syscalls	@ check range
+#ifdef CONFIG_CPU_SPECTRE
+		movhs	scno, #0
+		csdb
+#endif
 		stmloia	sp, {r5, r6}		@ shuffle args
 		movlo	r0, r1
 		movlo	r1, r2
diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S
index 0f07579af472..773424843d6e 100644
--- a/arch/arm/kernel/entry-header.S
+++ b/arch/arm/kernel/entry-header.S
@@ -378,6 +378,31 @@
 #endif
 	.endm
 
+	.macro	invoke_syscall, table, nr, tmp, ret, reload=0
+#ifdef CONFIG_CPU_SPECTRE
+	mov	\tmp, \nr
+	cmp	\tmp, #NR_syscalls		@ check upper syscall limit
+	movcs	\tmp, #0
+	csdb
+	badr	lr, \ret			@ return address
+	.if	\reload
+	add	r1, sp, #S_R0 + S_OFF		@ pointer to regs
+	ldmccia	r1, {r0 - r6}			@ reload r0-r6
+	stmccia	sp, {r4, r5}			@ update stack arguments
+	.endif
+	ldrcc	pc, [\table, \tmp, lsl #2]	@ call sys_* routine
+#else
+	cmp	\nr, #NR_syscalls		@ check upper syscall limit
+	badr	lr, \ret			@ return address
+	.if	\reload
+	add	r1, sp, #S_R0 + S_OFF		@ pointer to regs
+	ldmccia	r1, {r0 - r6}			@ reload r0-r6
+	stmccia	sp, {r4, r5}			@ update stack arguments
+	.endif
+	ldrcc	pc, [\table, \nr, lsl #2]	@ call sys_* routine
+#endif
+	.endm
+
 /*
  * These are the registers used in the syscall handler, and allow us to
  * have in theory up to 7 arguments to a function - r0 to r6.
diff --git a/arch/arm/kernel/head-nommu.S b/arch/arm/kernel/head-nommu.S
index 2e38f85b757a..dd546d65a383 100644
--- a/arch/arm/kernel/head-nommu.S
+++ b/arch/arm/kernel/head-nommu.S
@@ -68,14 +68,6 @@ ENTRY(stext)
 	beq	__error_p				@ yes, error 'p'
 
 #ifdef CONFIG_ARM_MPU
-	/* Calculate the size of a region covering just the kernel */
-	ldr	r5, =PLAT_PHYS_OFFSET		@ Region start: PHYS_OFFSET
-	ldr     r6, =(_end)			@ Cover whole kernel
-	sub	r6, r6, r5			@ Minimum size of region to map
-	clz	r6, r6				@ Region size must be 2^N...
-	rsb	r6, r6, #31			@ ...so round up region size
-	lsl	r6, r6, #MPU_RSR_SZ		@ Put size in right field
-	orr	r6, r6, #(1 << MPU_RSR_EN)	@ Set region enabled bit
 	bl	__setup_mpu
 #endif
 
@@ -83,8 +75,8 @@ ENTRY(stext)
 	ldr	r12, [r10, #PROCINFO_INITFUNC]
 	add	r12, r12, r10
 	ret	r12
-1:	bl	__after_proc_init
-	b	__mmap_switched
+1:	ldr	lr, =__mmap_switched
+	b	__after_proc_init
 ENDPROC(stext)
 
 #ifdef CONFIG_SMP
@@ -110,8 +102,6 @@ ENTRY(secondary_startup)
 	ldr	r7, __secondary_data
 
 #ifdef CONFIG_ARM_MPU
-	/* Use MPU region info supplied by __cpu_up */
-	ldr	r6, [r7]			@ get secondary_data.mpu_rgn_info
 	bl      __secondary_setup_mpu		@ Initialize the MPU
 #endif
 
@@ -133,12 +123,45 @@ __secondary_data:
 /*
  * Set the Control Register and Read the process ID.
  */
+	.text
 __after_proc_init:
+#ifdef CONFIG_ARM_MPU
+M_CLASS(movw	r12, #:lower16:BASEADDR_V7M_SCB)
+M_CLASS(movt	r12, #:upper16:BASEADDR_V7M_SCB)
+M_CLASS(ldr	r3, [r12, 0x50])
+AR_CLASS(mrc	p15, 0, r3, c0, c1, 4)          @ Read ID_MMFR0
+	and	r3, r3, #(MMFR0_PMSA)           @ PMSA field
+	teq	r3, #(MMFR0_PMSAv7)             @ PMSA v7
+	beq	1f
+	teq	r3, #(MMFR0_PMSAv8)		@ PMSA v8
+	/*
+	 * Memory region attributes for PMSAv8:
+	 *
+	 *   n = AttrIndx[2:0]
+	 *                      n       MAIR
+	 *   DEVICE_nGnRnE      000     00000000
+	 *   NORMAL             001     11111111
+	 */
+	ldreq	r3, =PMSAv8_MAIR(0x00, PMSAv8_RGN_DEVICE_nGnRnE) | \
+		     PMSAv8_MAIR(0xff, PMSAv8_RGN_NORMAL)
+AR_CLASS(mcreq	p15, 0, r3, c10, c2, 0)		@ MAIR 0
+M_CLASS(streq	r3, [r12, #PMSAv8_MAIR0])
+	moveq	r3, #0
+AR_CLASS(mcreq	p15, 0, r3, c10, c2, 1)		@ MAIR 1
+M_CLASS(streq	r3, [r12, #PMSAv8_MAIR1])
+
+1:
+#endif
 #ifdef CONFIG_CPU_CP15
 	/*
 	 * CP15 system control register value returned in r0 from
 	 * the CPU init function.
 	 */
+
+#ifdef CONFIG_ARM_MPU
+	biceq	r0, r0, #CR_BR			@ Disable the 'default mem-map'
+	orreq	r0, r0, #CR_M			@ Set SCTRL.M (MPU on)
+#endif
 #if defined(CONFIG_ALIGNMENT_TRAP) && __LINUX_ARM_ARCH__ < 6
 	orr	r0, r0, #CR_A
 #else
@@ -154,7 +177,15 @@ __after_proc_init:
 	bic	r0, r0, #CR_I
 #endif
 	mcr	p15, 0, r0, c1, c0, 0		@ write control reg
+	isb
 #elif defined (CONFIG_CPU_V7M)
+#ifdef CONFIG_ARM_MPU
+	ldreq	r3, [r12, MPU_CTRL]
+	biceq	r3, #MPU_CTRL_PRIVDEFENA
+	orreq	r3, #MPU_CTRL_ENABLE
+	streq	r3, [r12, MPU_CTRL]
+	isb
+#endif
 	/* For V7M systems we want to modify the CCR similarly to the SCTLR */
 #ifdef CONFIG_CPU_DCACHE_DISABLE
 	bic	r0, r0, #V7M_SCB_CCR_DC
@@ -165,9 +196,7 @@ __after_proc_init:
 #ifdef CONFIG_CPU_ICACHE_DISABLE
 	bic	r0, r0, #V7M_SCB_CCR_IC
 #endif
-	movw	r3, #:lower16:(BASEADDR_V7M_SCB + V7M_SCB_CCR)
-	movt	r3, #:upper16:(BASEADDR_V7M_SCB + V7M_SCB_CCR)
-	str	r0, [r3]
+	str	r0, [r12, V7M_SCB_CCR]
 #endif /* CONFIG_CPU_CP15 elif CONFIG_CPU_V7M */
 	ret	lr
 ENDPROC(__after_proc_init)
@@ -184,7 +213,7 @@ ENDPROC(__after_proc_init)
 .endm
 
 /* Setup a single MPU region, either D or I side (D-side for unified) */
-.macro setup_region bar, acr, sr, side = MPU_DATA_SIDE, unused
+.macro setup_region bar, acr, sr, side = PMSAv7_DATA_SIDE, unused
 	mcr	p15, 0, \bar, c6, c1, (0 + \side)	@ I/DRBAR
 	mcr	p15, 0, \acr, c6, c1, (4 + \side)	@ I/DRACR
 	mcr	p15, 0, \sr, c6, c1, (2 + \side)		@ I/DRSR
@@ -192,14 +221,14 @@ ENDPROC(__after_proc_init)
 #else
 .macro set_region_nr tmp, rgnr, base
 	mov	\tmp, \rgnr
-	str     \tmp, [\base, #MPU_RNR]
+	str     \tmp, [\base, #PMSAv7_RNR]
 .endm
 
 .macro setup_region bar, acr, sr, unused, base
 	lsl     \acr, \acr, #16
 	orr     \acr, \acr, \sr
-	str     \bar, [\base, #MPU_RBAR]
-	str     \acr, [\base, #MPU_RASR]
+	str     \bar, [\base, #PMSAv7_RBAR]
+	str     \acr, [\base, #PMSAv7_RASR]
 .endm
 
 #endif
@@ -210,8 +239,9 @@ ENDPROC(__after_proc_init)
  * Region 2: Normal, Shared, cacheable for RAM. From PHYS_OFFSET, size from r6
  * Region 3: Normal, shared, inaccessible from PL0 to protect the vectors page
  *
- * r6: Value to be written to DRSR (and IRSR if required) for MPU_RAM_REGION
+ * r6: Value to be written to DRSR (and IRSR if required) for PMSAv7_RAM_REGION
 */
+	__HEAD
 
 ENTRY(__setup_mpu)
 
@@ -223,7 +253,22 @@ AR_CLASS(mrc	p15, 0, r0, c0, c1, 4)		@ Read ID_MMFR0
 M_CLASS(ldr	r0, [r12, 0x50])
 	and	r0, r0, #(MMFR0_PMSA)		@ PMSA field
 	teq	r0, #(MMFR0_PMSAv7)		@ PMSA v7
-	bxne	lr
+	beq	__setup_pmsa_v7
+	teq	r0, #(MMFR0_PMSAv8)		@ PMSA v8
+	beq	__setup_pmsa_v8
+
+	ret	lr
+ENDPROC(__setup_mpu)
+
+ENTRY(__setup_pmsa_v7)
+	/* Calculate the size of a region covering just the kernel */
+	ldr	r5, =PLAT_PHYS_OFFSET		@ Region start: PHYS_OFFSET
+	ldr     r6, =(_end)			@ Cover whole kernel
+	sub	r6, r6, r5			@ Minimum size of region to map
+	clz	r6, r6				@ Region size must be 2^N...
+	rsb	r6, r6, #31			@ ...so round up region size
+	lsl	r6, r6, #PMSAv7_RSR_SZ		@ Put size in right field
+	orr	r6, r6, #(1 << PMSAv7_RSR_EN)	@ Set region enabled bit
 
 	/* Determine whether the D/I-side memory map is unified. We set the
 	 * flags here and continue to use them for the rest of this function */
@@ -234,77 +279,189 @@ M_CLASS(ldr    r0, [r12, #MPU_TYPE])
 	tst	r0, #MPUIR_nU			@ MPUIR_nU = 0 for unified
 
 	/* Setup second region first to free up r6 */
-	set_region_nr r0, #MPU_RAM_REGION, r12
+	set_region_nr r0, #PMSAv7_RAM_REGION, r12
 	isb
 	/* Full access from PL0, PL1, shared for CONFIG_SMP, cacheable */
 	ldr	r0, =PLAT_PHYS_OFFSET		@ RAM starts at PHYS_OFFSET
-	ldr	r5,=(MPU_AP_PL1RW_PL0RW | MPU_RGN_NORMAL)
+	ldr	r5,=(PMSAv7_AP_PL1RW_PL0RW | PMSAv7_RGN_NORMAL)
 
-	setup_region r0, r5, r6, MPU_DATA_SIDE, r12	@ PHYS_OFFSET, shared, enabled
+	setup_region r0, r5, r6, PMSAv7_DATA_SIDE, r12	@ PHYS_OFFSET, shared, enabled
 	beq	1f					@ Memory-map not unified
-	setup_region r0, r5, r6, MPU_INSTR_SIDE, r12	@ PHYS_OFFSET, shared, enabled
+	setup_region r0, r5, r6, PMSAv7_INSTR_SIDE, r12	@ PHYS_OFFSET, shared, enabled
 1:	isb
 
 	/* First/background region */
-	set_region_nr r0, #MPU_BG_REGION, r12
+	set_region_nr r0, #PMSAv7_BG_REGION, r12
 	isb
 	/* Execute Never,  strongly ordered, inaccessible to PL0, rw PL1  */
 	mov	r0, #0				@ BG region starts at 0x0
-	ldr	r5,=(MPU_ACR_XN | MPU_RGN_STRONGLY_ORDERED | MPU_AP_PL1RW_PL0NA)
-	mov	r6, #MPU_RSR_ALL_MEM		@ 4GB region, enabled
+	ldr	r5,=(PMSAv7_ACR_XN | PMSAv7_RGN_STRONGLY_ORDERED | PMSAv7_AP_PL1RW_PL0NA)
+	mov	r6, #PMSAv7_RSR_ALL_MEM		@ 4GB region, enabled
 
-	setup_region r0, r5, r6, MPU_DATA_SIDE, r12	@ 0x0, BG region, enabled
+	setup_region r0, r5, r6, PMSAv7_DATA_SIDE, r12	@ 0x0, BG region, enabled
 	beq	2f					@ Memory-map not unified
-	setup_region r0, r5, r6, MPU_INSTR_SIDE r12	@ 0x0, BG region, enabled
+	setup_region r0, r5, r6, PMSAv7_INSTR_SIDE r12	@ 0x0, BG region, enabled
 2:	isb
 
 #ifdef CONFIG_XIP_KERNEL
-	set_region_nr r0, #MPU_ROM_REGION, r12
+	set_region_nr r0, #PMSAv7_ROM_REGION, r12
 	isb
 
-	ldr	r5,=(MPU_AP_PL1RO_PL0NA | MPU_RGN_NORMAL)
+	ldr	r5,=(PMSAv7_AP_PL1RO_PL0NA | PMSAv7_RGN_NORMAL)
 
 	ldr	r0, =CONFIG_XIP_PHYS_ADDR		@ ROM start
 	ldr     r6, =(_exiprom)				@ ROM end
 	sub	r6, r6, r0				@ Minimum size of region to map
 	clz	r6, r6					@ Region size must be 2^N...
 	rsb	r6, r6, #31				@ ...so round up region size
-	lsl	r6, r6, #MPU_RSR_SZ			@ Put size in right field
-	orr	r6, r6, #(1 << MPU_RSR_EN)		@ Set region enabled bit
+	lsl	r6, r6, #PMSAv7_RSR_SZ			@ Put size in right field
+	orr	r6, r6, #(1 << PMSAv7_RSR_EN)		@ Set region enabled bit
 
-	setup_region r0, r5, r6, MPU_DATA_SIDE, r12	@ XIP_PHYS_ADDR, shared, enabled
+	setup_region r0, r5, r6, PMSAv7_DATA_SIDE, r12	@ XIP_PHYS_ADDR, shared, enabled
 	beq	3f					@ Memory-map not unified
-	setup_region r0, r5, r6, MPU_INSTR_SIDE, r12	@ XIP_PHYS_ADDR, shared, enabled
+	setup_region r0, r5, r6, PMSAv7_INSTR_SIDE, r12	@ XIP_PHYS_ADDR, shared, enabled
 3:	isb
 #endif
+	ret	lr
+ENDPROC(__setup_pmsa_v7)
+
+ENTRY(__setup_pmsa_v8)
+	mov	r0, #0
+AR_CLASS(mcr	p15, 0, r0, c6, c2, 1)		@ PRSEL
+M_CLASS(str	r0, [r12, #PMSAv8_RNR])
+	isb
+
+#ifdef CONFIG_XIP_KERNEL
+	ldr	r5, =CONFIG_XIP_PHYS_ADDR		@ ROM start
+	ldr     r6, =(_exiprom)				@ ROM end
+	sub	r6, r6, #1
+	bic	r6, r6, #(PMSAv8_MINALIGN - 1)
+
+	orr	r5, r5, #(PMSAv8_AP_PL1RW_PL0NA | PMSAv8_RGN_SHARED)
+	orr	r6, r6, #(PMSAv8_LAR_IDX(PMSAv8_RGN_NORMAL) | PMSAv8_LAR_EN)
+
+AR_CLASS(mcr	p15, 0, r5, c6, c8, 0)			@ PRBAR0
+AR_CLASS(mcr	p15, 0, r6, c6, c8, 1)			@ PRLAR0
+M_CLASS(str	r5, [r12, #PMSAv8_RBAR_A(0)])
+M_CLASS(str	r6, [r12, #PMSAv8_RLAR_A(0)])
+#endif
+
+	ldr	r5, =KERNEL_START
+	ldr	r6, =KERNEL_END
+	sub	r6, r6, #1
+	bic	r6, r6, #(PMSAv8_MINALIGN - 1)
+
+	orr	r5, r5, #(PMSAv8_AP_PL1RW_PL0NA | PMSAv8_RGN_SHARED)
+	orr	r6, r6, #(PMSAv8_LAR_IDX(PMSAv8_RGN_NORMAL) | PMSAv8_LAR_EN)
+
+AR_CLASS(mcr	p15, 0, r5, c6, c8, 4)			@ PRBAR1
+AR_CLASS(mcr	p15, 0, r6, c6, c8, 5)			@ PRLAR1
+M_CLASS(str	r5, [r12, #PMSAv8_RBAR_A(1)])
+M_CLASS(str	r6, [r12, #PMSAv8_RLAR_A(1)])
+
+	/* Setup Background: 0x0 - min(KERNEL_START, XIP_PHYS_ADDR) */
+#ifdef CONFIG_XIP_KERNEL
+	ldr	r6, =KERNEL_START
+	ldr	r5, =CONFIG_XIP_PHYS_ADDR
+	cmp	r6, r5
+	movcs	r6, r5
+#else
+	ldr	r6, =KERNEL_START
+#endif
+	cmp	r6, #0
+	beq	1f
+
+	mov	r5, #0
+	sub	r6, r6, #1
+	bic	r6, r6, #(PMSAv8_MINALIGN - 1)
+
+	orr	r5, r5, #(PMSAv8_AP_PL1RW_PL0NA | PMSAv8_RGN_SHARED | PMSAv8_BAR_XN)
+	orr	r6, r6, #(PMSAv8_LAR_IDX(PMSAv8_RGN_DEVICE_nGnRnE) | PMSAv8_LAR_EN)
+
+AR_CLASS(mcr	p15, 0, r5, c6, c9, 0)			@ PRBAR2
+AR_CLASS(mcr	p15, 0, r6, c6, c9, 1)			@ PRLAR2
+M_CLASS(str	r5, [r12, #PMSAv8_RBAR_A(2)])
+M_CLASS(str	r6, [r12, #PMSAv8_RLAR_A(2)])
+
+1:
+	/* Setup Background: max(KERNEL_END, _exiprom) - 0xffffffff */
+#ifdef CONFIG_XIP_KERNEL
+	ldr	r5, =KERNEL_END
+	ldr	r6, =(_exiprom)
+	cmp	r5, r6
+	movcc	r5, r6
+#else
+	ldr	r5, =KERNEL_END
+#endif
+	mov	r6, #0xffffffff
+	bic	r6, r6, #(PMSAv8_MINALIGN - 1)
+
+	orr	r5, r5, #(PMSAv8_AP_PL1RW_PL0NA | PMSAv8_RGN_SHARED | PMSAv8_BAR_XN)
+	orr	r6, r6, #(PMSAv8_LAR_IDX(PMSAv8_RGN_DEVICE_nGnRnE) | PMSAv8_LAR_EN)
 
-	/* Enable the MPU */
-AR_CLASS(mrc	p15, 0, r0, c1, c0, 0)		@ Read SCTLR
-AR_CLASS(bic	r0, r0, #CR_BR)			@ Disable the 'default mem-map'
-AR_CLASS(orr	r0, r0, #CR_M)			@ Set SCTRL.M (MPU on)
-AR_CLASS(mcr	p15, 0, r0, c1, c0, 0)		@ Enable MPU
+AR_CLASS(mcr	p15, 0, r5, c6, c9, 4)			@ PRBAR3
+AR_CLASS(mcr	p15, 0, r6, c6, c9, 5)			@ PRLAR3
+M_CLASS(str	r5, [r12, #PMSAv8_RBAR_A(3)])
+M_CLASS(str	r6, [r12, #PMSAv8_RLAR_A(3)])
 
-M_CLASS(ldr	r0, [r12, #MPU_CTRL])
-M_CLASS(bic	r0, #MPU_CTRL_PRIVDEFENA)
-M_CLASS(orr	r0, #MPU_CTRL_ENABLE)
-M_CLASS(str	r0, [r12, #MPU_CTRL])
+#ifdef CONFIG_XIP_KERNEL
+	/* Setup Background: min(_exiprom, KERNEL_END) - max(KERNEL_START, XIP_PHYS_ADDR) */
+	ldr	r5, =(_exiprom)
+	ldr	r6, =KERNEL_END
+	cmp	r5, r6
+	movcs	r5, r6
+
+	ldr	r6, =KERNEL_START
+	ldr	r0, =CONFIG_XIP_PHYS_ADDR
+	cmp	r6, r0
+	movcc	r6, r0
+
+	sub	r6, r6, #1
+	bic	r6, r6, #(PMSAv8_MINALIGN - 1)
+
+	orr	r5, r5, #(PMSAv8_AP_PL1RW_PL0NA | PMSAv8_RGN_SHARED | PMSAv8_BAR_XN)
+	orr	r6, r6, #(PMSAv8_LAR_IDX(PMSAv8_RGN_DEVICE_nGnRnE) | PMSAv8_LAR_EN)
+
+#ifdef CONFIG_CPU_V7M
+	/* There is no alias for n == 4 */
+	mov	r0, #4
+	str	r0, [r12, #PMSAv8_RNR]			@ PRSEL
 	isb
 
+	str	r5, [r12, #PMSAv8_RBAR_A(0)]
+	str	r6, [r12, #PMSAv8_RLAR_A(0)]
+#else
+	mcr	p15, 0, r5, c6, c10, 1			@ PRBAR4
+	mcr	p15, 0, r6, c6, c10, 2			@ PRLAR4
+#endif
+#endif
 	ret	lr
-ENDPROC(__setup_mpu)
+ENDPROC(__setup_pmsa_v8)
 
 #ifdef CONFIG_SMP
 /*
  * r6: pointer at mpu_rgn_info
  */
 
+	.text
 ENTRY(__secondary_setup_mpu)
+	/* Use MPU region info supplied by __cpu_up */
+	ldr	r6, [r7]			@ get secondary_data.mpu_rgn_info
+
 	/* Probe for v7 PMSA compliance */
 	mrc	p15, 0, r0, c0, c1, 4		@ Read ID_MMFR0
 	and	r0, r0, #(MMFR0_PMSA)		@ PMSA field
 	teq	r0, #(MMFR0_PMSAv7)		@ PMSA v7
-	bne	__error_p
+	beq	__secondary_setup_pmsa_v7
+	teq	r0, #(MMFR0_PMSAv8)		@ PMSA v8
+	beq	__secondary_setup_pmsa_v8
+	b	__error_p
+ENDPROC(__secondary_setup_mpu)
 
+/*
+ * r6: pointer at mpu_rgn_info
+ */
+ENTRY(__secondary_setup_pmsa_v7)
 	/* Determine whether the D/I-side memory map is unified. We set the
 	 * flags here and continue to use them for the rest of this function */
 	mrc	p15, 0, r0, c0, c0, 4		@ MPUIR
@@ -328,25 +485,45 @@ ENTRY(__secondary_setup_mpu)
 	ldr	r6, [r3, #MPU_RGN_DRSR]
 	ldr	r5, [r3, #MPU_RGN_DRACR]
 
-	setup_region r0, r5, r6, MPU_DATA_SIDE
+	setup_region r0, r5, r6, PMSAv7_DATA_SIDE
 	beq	2f
-	setup_region r0, r5, r6, MPU_INSTR_SIDE
+	setup_region r0, r5, r6, PMSAv7_INSTR_SIDE
 2:	isb
 
 	mrc	p15, 0, r0, c0, c0, 4		@ Reevaluate the MPUIR
 	cmp	r4, #0
 	bgt	1b
 
-	/* Enable the MPU */
-	mrc	p15, 0, r0, c1, c0, 0		@ Read SCTLR
-	bic	r0, r0, #CR_BR			@ Disable the 'default mem-map'
-	orr	r0, r0, #CR_M			@ Set SCTRL.M (MPU on)
-	mcr	p15, 0, r0, c1, c0, 0		@ Enable MPU
+	ret	lr
+ENDPROC(__secondary_setup_pmsa_v7)
+
+ENTRY(__secondary_setup_pmsa_v8)
+	ldr	r4, [r6, #MPU_RNG_INFO_USED]
+#ifndef CONFIG_XIP_KERNEL
+	add	r4, r4, #1
+#endif
+	mov	r5, #MPU_RNG_SIZE
+	add	r3, r6, #MPU_RNG_INFO_RNGS
+	mla	r3, r4, r5, r3
+
+1:
+	sub	r3, r3, #MPU_RNG_SIZE
+	sub	r4, r4, #1
+
+	mcr	p15, 0, r4, c6, c2, 1		@ PRSEL
 	isb
 
-	ret	lr
-ENDPROC(__secondary_setup_mpu)
+	ldr	r5, [r3, #MPU_RGN_PRBAR]
+	ldr	r6, [r3, #MPU_RGN_PRLAR]
 
+	mcr	p15, 0, r5, c6, c3, 0		@ PRBAR
+	mcr	p15, 0, r6, c6, c3, 1           @ PRLAR
+
+	cmp	r4, #0
+	bgt	1b
+
+	ret	lr
+ENDPROC(__secondary_setup_pmsa_v8)
 #endif /* CONFIG_SMP */
 #endif /* CONFIG_ARM_MPU */
 #include "head-common.S"
diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
index 2da087926ebe..0978282d5fc2 100644
--- a/arch/arm/kernel/smp.c
+++ b/arch/arm/kernel/smp.c
@@ -31,6 +31,7 @@
 #include <linux/irq_work.h>
 
 #include <linux/atomic.h>
+#include <asm/bugs.h>
 #include <asm/smp.h>
 #include <asm/cacheflush.h>
 #include <asm/cpu.h>
@@ -236,8 +237,6 @@ int __cpu_disable(void)
 	flush_cache_louis();
 	local_flush_tlb_all();
 
-	clear_tasks_mm_cpumask(cpu);
-
 	return 0;
 }
 
@@ -255,6 +254,7 @@ void __cpu_die(unsigned int cpu)
 	}
 	pr_debug("CPU%u: shutdown\n", cpu);
 
+	clear_tasks_mm_cpumask(cpu);
 	/*
 	 * platform_cpu_kill() is generally expected to do the powering off
 	 * and/or cutting of clocks to the dying CPU.  Optionally, this may
@@ -405,6 +405,9 @@ asmlinkage void secondary_start_kernel(void)
 	 * before we continue - which happens after __cpu_up returns.
 	 */
 	set_cpu_online(cpu, true);
+
+	check_other_bugs();
+
 	complete(&cpu_running);
 
 	local_irq_enable();
diff --git a/arch/arm/kernel/suspend.c b/arch/arm/kernel/suspend.c
index a40ebb7c0896..d08099269e35 100644
--- a/arch/arm/kernel/suspend.c
+++ b/arch/arm/kernel/suspend.c
@@ -3,6 +3,7 @@
 #include <linux/slab.h>
 #include <linux/mm_types.h>
 
+#include <asm/bugs.h>
 #include <asm/cacheflush.h>
 #include <asm/idmap.h>
 #include <asm/pgalloc.h>
@@ -36,6 +37,7 @@ int cpu_suspend(unsigned long arg, int (*fn)(unsigned long))
 		cpu_switch_mm(mm->pgd, mm);
 		local_flush_bp_all();
 		local_flush_tlb_all();
+		check_other_bugs();
 	}
 
 	return ret;
diff --git a/arch/arm/kernel/vmlinux-xip.lds.S b/arch/arm/kernel/vmlinux-xip.lds.S
index d32f5d35f602..3593d5c1acd2 100644
--- a/arch/arm/kernel/vmlinux-xip.lds.S
+++ b/arch/arm/kernel/vmlinux-xip.lds.S
@@ -13,6 +13,7 @@
 #include <asm/cache.h>
 #include <asm/thread_info.h>
 #include <asm/memory.h>
+#include <asm/mpu.h>
 #include <asm/page.h>
 
 #include "vmlinux.lds.h"
@@ -148,6 +149,9 @@ SECTIONS
 	__init_end = .;
 
 	BSS_SECTION(0, 0, 8)
+#ifdef CONFIG_ARM_MPU
+	. = ALIGN(PMSAv8_MINALIGN);
+#endif
 	_end = .;
 
 	STABS_DEBUG
diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S
index b77dc675ae55..23150c0f0f4d 100644
--- a/arch/arm/kernel/vmlinux.lds.S
+++ b/arch/arm/kernel/vmlinux.lds.S
@@ -12,6 +12,7 @@
 #include <asm/cache.h>
 #include <asm/thread_info.h>
 #include <asm/memory.h>
+#include <asm/mpu.h>
 #include <asm/page.h>
 #include <asm/pgtable.h>
 
@@ -54,6 +55,9 @@ SECTIONS
 	. = ALIGN(1<<SECTION_SHIFT);
 #endif
 
+#ifdef CONFIG_ARM_MPU
+	. = ALIGN(PMSAv8_MINALIGN);
+#endif
 	.text : {			/* Real text segment		*/
 		_stext = .;		/* Text and read-only data	*/
 		ARM_TEXT
@@ -143,6 +147,9 @@ SECTIONS
 	_edata = .;
 
 	BSS_SECTION(0, 0, 0)
+#ifdef CONFIG_ARM_MPU
+	. = ALIGN(PMSAv8_MINALIGN);
+#endif
 	_end = .;
 
 	STABS_DEBUG
diff --git a/arch/arm/kernel/vmlinux.lds.h b/arch/arm/kernel/vmlinux.lds.h
index 71281e08e1d4..ae5fdff18406 100644
--- a/arch/arm/kernel/vmlinux.lds.h
+++ b/arch/arm/kernel/vmlinux.lds.h
@@ -27,24 +27,24 @@
 
 #define PROC_INFO							\
 		. = ALIGN(4);						\
-		VMLINUX_SYMBOL(__proc_info_begin) = .;			\
+		__proc_info_begin = .;					\
 		*(.proc.info.init)					\
-		VMLINUX_SYMBOL(__proc_info_end) = .;
+		__proc_info_end = .;
 
 #define HYPERVISOR_TEXT							\
-		VMLINUX_SYMBOL(__hyp_text_start) = .;			\
+		__hyp_text_start = .;					\
 		*(.hyp.text)						\
-		VMLINUX_SYMBOL(__hyp_text_end) = .;
+		__hyp_text_end = .;
 
 #define IDMAP_TEXT							\
 		ALIGN_FUNCTION();					\
-		VMLINUX_SYMBOL(__idmap_text_start) = .;			\
+		__idmap_text_start = .;					\
 		*(.idmap.text)						\
-		VMLINUX_SYMBOL(__idmap_text_end) = .;			\
+		__idmap_text_end = .;					\
 		. = ALIGN(PAGE_SIZE);					\
-		VMLINUX_SYMBOL(__hyp_idmap_text_start) = .;		\
+		__hyp_idmap_text_start = .;				\
 		*(.hyp.idmap.text)					\
-		VMLINUX_SYMBOL(__hyp_idmap_text_end) = .;
+		__hyp_idmap_text_end = .;
 
 #define ARM_DISCARD							\
 		*(.ARM.exidx.exit.text)					\