Enable v8.6 AMU enhancements (FEAT_AMUv1p1)

ARMv8.6 adds virtual offset registers to support virtualization of the
event counters in EL1 and EL0.  This patch enables support for this
feature in EL3 firmware.
Signed-off-by: John Powell <john.powell@arm.com>
Change-Id: I7ee1f3d9f554930bf5ef6f3d492e932e6d95b217

Makefile

@@ -895,6 +895,7 @@ $(eval $(call assert_booleans,\
         DYN_DISABLE_AUTH \
         EL3_EXCEPTION_HANDLING \
         ENABLE_AMU \
+        AMU_RESTRICT_COUNTERS \
         ENABLE_ASSERTIONS \
         ENABLE_MPAM_FOR_LOWER_ELS \
         ENABLE_PIE \
@@ -984,6 +985,7 @@ $(eval $(call add_defines,\
         DECRYPTION_SUPPORT_${DECRYPTION_SUPPORT} \
         DISABLE_MTPMU \
         ENABLE_AMU \
+        AMU_RESTRICT_COUNTERS \
         ENABLE_ASSERTIONS \
         ENABLE_BTI \
         ENABLE_MPAM_FOR_LOWER_ELS \

docs/getting_started/build-options.rst

@@ -22,6 +22,10 @@ Common build options
    directory containing the SP source, relative to the ``bl32/``; the directory
    is expected to contain a makefile called ``<aarch32_sp-value>.mk``.
 
+-  ``AMU_RESTRICT_COUNTERS``: Register reads to the group 1 counters will return
+   zero at all but the highest implemented exception level.  Reads from the
+   memory mapped view are unaffected by this control.
+
 -  ``ARCH`` : Choose the target build architecture for TF-A. It can take either
    ``aarch64`` or ``aarch32`` as values. By default, it is defined to
    ``aarch64``.

include/arch/aarch32/arch.h

 /*
- * Copyright (c) 2016-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2016-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -116,6 +116,9 @@
 #define ID_PFR0_AMU_SHIFT	U(20)
 #define ID_PFR0_AMU_LENGTH	U(4)
 #define ID_PFR0_AMU_MASK	U(0xf)
+#define ID_PFR0_AMU_NOT_SUPPORTED	U(0x0)
+#define ID_PFR0_AMU_V1		U(0x1)
+#define ID_PFR0_AMU_V1P1	U(0x2)
 
 #define ID_PFR0_DIT_SHIFT	U(24)
 #define ID_PFR0_DIT_LENGTH	U(4)
@@ -653,7 +656,7 @@
 #define PAR_ADDR_MASK	(BIT_64(40) - ULL(1)) /* 40-bits-wide page address */
 
 /*******************************************************************************
- * Definitions for system register interface to AMU for ARMv8.4 onwards
+ * Definitions for system register interface to AMU for FEAT_AMUv1
  ******************************************************************************/
 #define AMCR		p15, 0, c13, c2, 0
 #define AMCFGR		p15, 0, c13, c2, 1
@@ -712,6 +715,9 @@
 #define AMEVTYPER1E	p15, 0, c13, c15, 6
 #define AMEVTYPER1F	p15, 0, c13, c15, 7
 
+/* AMCR definitions */
+#define AMCR_CG1RZ_BIT		(ULL(1) << 17)
+
 /* AMCFGR definitions */
 #define AMCFGR_NCG_SHIFT	U(28)
 #define AMCFGR_NCG_MASK		U(0xf)

include/arch/aarch32/arch_helpers.h

 /*
- * Copyright (c) 2016-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2016-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -303,6 +303,7 @@ DEFINE_COPROCR_RW_FUNCS(dacr, DACR)
 /* Coproc registers for 32bit AMU support */
 DEFINE_COPROCR_READ_FUNC(amcfgr, AMCFGR)
 DEFINE_COPROCR_READ_FUNC(amcgcr, AMCGCR)
+DEFINE_COPROCR_RW_FUNCS(amcr, AMCR)
 
 DEFINE_COPROCR_RW_FUNCS(amcntenset0, AMCNTENSET0)
 DEFINE_COPROCR_RW_FUNCS(amcntenset1, AMCNTENSET1)

include/arch/aarch64/arch.h

 /*
- * Copyright (c) 2013-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2013-2021, ARM Limited and Contributors. All rights reserved.
  * Copyright (c) 2020, NVIDIA Corporation. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
@@ -161,6 +161,9 @@
 #define ID_AA64PFR0_EL3_SHIFT	U(12)
 #define ID_AA64PFR0_AMU_SHIFT	U(44)
 #define ID_AA64PFR0_AMU_MASK	ULL(0xf)
+#define ID_AA64PFR0_AMU_NOT_SUPPORTED	U(0x0)
+#define ID_AA64PFR0_AMU_V1	U(0x1)
+#define ID_AA64PFR0_AMU_V1P1	U(0x2)
 #define ID_AA64PFR0_ELX_MASK	ULL(0xf)
 #define ID_AA64PFR0_GIC_SHIFT	U(24)
 #define ID_AA64PFR0_GIC_WIDTH	U(4)
@@ -406,9 +409,10 @@
 #define SCR_RES1_BITS		((U(1) << 4) | (U(1) << 5))
 #define SCR_TWEDEL_SHIFT	U(30)
 #define SCR_TWEDEL_MASK		ULL(0xf)
+#define SCR_AMVOFFEN_BIT	(UL(1) << 35)
 #define SCR_TWEDEn_BIT		(UL(1) << 29)
-#define SCR_ECVEN_BIT           (UL(1) << 28)
-#define SCR_FGTEN_BIT           (UL(1) << 27)
+#define SCR_ECVEN_BIT		(UL(1) << 28)
+#define SCR_FGTEN_BIT		(UL(1) << 27)
 #define SCR_ATA_BIT		(UL(1) << 26)
 #define SCR_FIEN_BIT		(UL(1) << 21)
 #define SCR_EEL2_BIT		(UL(1) << 18)
@@ -479,6 +483,7 @@
 #define VTTBR_BADDR_SHIFT	U(0)
 
 /* HCR definitions */
+#define HCR_AMVOFFEN_BIT	(ULL(1) << 51)
 #define HCR_API_BIT		(ULL(1) << 41)
 #define HCR_APK_BIT		(ULL(1) << 40)
 #define HCR_E2H_BIT		(ULL(1) << 34)
@@ -721,13 +726,13 @@
 #define MAX_CACHE_LINE_SIZE	U(0x800) /* 2KB */
 
 /* Physical timer control register bit fields shifts and masks */
-#define CNTP_CTL_ENABLE_SHIFT   U(0)
-#define CNTP_CTL_IMASK_SHIFT    U(1)
-#define CNTP_CTL_ISTATUS_SHIFT  U(2)
+#define CNTP_CTL_ENABLE_SHIFT	U(0)
+#define CNTP_CTL_IMASK_SHIFT	U(1)
+#define CNTP_CTL_ISTATUS_SHIFT	U(2)
 
-#define CNTP_CTL_ENABLE_MASK    U(1)
-#define CNTP_CTL_IMASK_MASK     U(1)
-#define CNTP_CTL_ISTATUS_MASK   U(1)
+#define CNTP_CTL_ENABLE_MASK	U(1)
+#define CNTP_CTL_IMASK_MASK	U(1)
+#define CNTP_CTL_ISTATUS_MASK	U(1)
 
 /* Physical timer control macros */
 #define CNTP_CTL_ENABLE_BIT	(U(1) << CNTP_CTL_ENABLE_SHIFT)
@@ -913,7 +918,7 @@
 #define MPAM3_EL3		S3_6_C10_C5_0
 
 /*******************************************************************************
- * Definitions for system register interface to AMU for ARMv8.4 onwards
+ * Definitions for system register interface to AMU for FEAT_AMUv1
  ******************************************************************************/
 #define AMCR_EL0		S3_3_C13_C2_0
 #define AMCFGR_EL0		S3_3_C13_C2_1
@@ -991,6 +996,50 @@
 
 #define MPAMIDR_HAS_HCR_BIT		(ULL(1) << 17)
 
+/*******************************************************************************
+ * Definitions for system register interface to AMU for FEAT_AMUv1p1
+ ******************************************************************************/
+
+/* Definition for register defining which virtual offsets are implemented. */
+#define AMCG1IDR_EL0		S3_3_C13_C2_6
+#define AMCG1IDR_CTR_MASK	ULL(0xffff)
+#define AMCG1IDR_CTR_SHIFT	U(0)
+#define AMCG1IDR_VOFF_MASK	ULL(0xffff)
+#define AMCG1IDR_VOFF_SHIFT	U(16)
+
+/* New bit added to AMCR_EL0 */
+#define AMCR_CG1RZ_BIT		(ULL(0x1) << 17)
+
+/*
+ * Definitions for virtual offset registers for architected activity monitor
+ * event counters.
+ * AMEVCNTVOFF01_EL2 intentionally left undefined, as it does not exist.
+ */
+#define AMEVCNTVOFF00_EL2	S3_4_C13_C8_0
+#define AMEVCNTVOFF02_EL2	S3_4_C13_C8_2
+#define AMEVCNTVOFF03_EL2	S3_4_C13_C8_3
+
+/*
+ * Definitions for virtual offset registers for auxiliary activity monitor event
+ * counters.
+ */
+#define AMEVCNTVOFF10_EL2	S3_4_C13_C10_0
+#define AMEVCNTVOFF11_EL2	S3_4_C13_C10_1
+#define AMEVCNTVOFF12_EL2	S3_4_C13_C10_2
+#define AMEVCNTVOFF13_EL2	S3_4_C13_C10_3
+#define AMEVCNTVOFF14_EL2	S3_4_C13_C10_4
+#define AMEVCNTVOFF15_EL2	S3_4_C13_C10_5
+#define AMEVCNTVOFF16_EL2	S3_4_C13_C10_6
+#define AMEVCNTVOFF17_EL2	S3_4_C13_C10_7
+#define AMEVCNTVOFF18_EL2	S3_4_C13_C11_0
+#define AMEVCNTVOFF19_EL2	S3_4_C13_C11_1
+#define AMEVCNTVOFF1A_EL2	S3_4_C13_C11_2
+#define AMEVCNTVOFF1B_EL2	S3_4_C13_C11_3
+#define AMEVCNTVOFF1C_EL2	S3_4_C13_C11_4
+#define AMEVCNTVOFF1D_EL2	S3_4_C13_C11_5
+#define AMEVCNTVOFF1E_EL2	S3_4_C13_C11_6
+#define AMEVCNTVOFF1F_EL2	S3_4_C13_C11_7
+
 /*******************************************************************************
  * RAS system registers
  ******************************************************************************/

include/arch/aarch64/arch_features.h

 /*
- * Copyright (c) 2019-2020, Arm Limited. All rights reserved.
+ * Copyright (c) 2019-2021, Arm Limited. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -82,6 +82,12 @@ static inline bool is_armv8_5_rng_present(void)
 		ID_AA64ISAR0_RNDR_MASK);
 }
 
+static inline bool is_armv8_6_feat_amuv1p1_present(void)
+{
+	return (((read_id_aa64pfr0_el1() >> ID_AA64PFR0_AMU_SHIFT) &
+		ID_AA64PFR0_AMU_MASK) >= ID_AA64PFR0_AMU_V1P1);
+}
+
 /*
  * Return MPAM version:
  *

include/arch/aarch64/arch_helpers.h

 /*
- * Copyright (c) 2013-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2013-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -485,6 +485,8 @@ DEFINE_RENAME_SYSREG_RW_FUNCS(icc_sgi1r, ICC_SGI1R)
 
 DEFINE_RENAME_SYSREG_READ_FUNC(amcfgr_el0, AMCFGR_EL0)
 DEFINE_RENAME_SYSREG_READ_FUNC(amcgcr_el0, AMCGCR_EL0)
+DEFINE_RENAME_SYSREG_READ_FUNC(amcg1idr_el0, AMCG1IDR_EL0)
+DEFINE_RENAME_SYSREG_RW_FUNCS(amcr_el0, AMCR_EL0)
 DEFINE_RENAME_SYSREG_RW_FUNCS(amcntenclr0_el0, AMCNTENCLR0_EL0)
 DEFINE_RENAME_SYSREG_RW_FUNCS(amcntenset0_el0, AMCNTENSET0_EL0)
 DEFINE_RENAME_SYSREG_RW_FUNCS(amcntenclr1_el0, AMCNTENCLR1_EL0)

include/lib/extensions/amu.h

 /*
- * Copyright (c) 2017-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2017-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -66,19 +66,31 @@ CASSERT(AMU_GROUP1_COUNTERS_MASK <= 0xffff, invalid_amu_group1_counters_mask);
 
 struct amu_ctx {
 	uint64_t group0_cnts[AMU_GROUP0_NR_COUNTERS];
+#if __aarch64__
+	/* Architected event counter 1 does not have an offset register. */
+	uint64_t group0_voffsets[AMU_GROUP0_NR_COUNTERS-1];
+#endif
 
 #if AMU_GROUP1_NR_COUNTERS
 	uint64_t group1_cnts[AMU_GROUP1_NR_COUNTERS];
+#if __aarch64__
+	uint64_t group1_voffsets[AMU_GROUP1_NR_COUNTERS];
+#endif
 #endif
 };
 
-bool amu_supported(void);
+unsigned int amu_get_version(void);
 void amu_enable(bool el2_unused);
 
 /* Group 0 configuration helpers */
 uint64_t amu_group0_cnt_read(unsigned int idx);
 void amu_group0_cnt_write(unsigned int idx, uint64_t val);
 
+#if __aarch64__
+uint64_t amu_group0_voffset_read(unsigned int idx);
+void amu_group0_voffset_write(unsigned int idx, uint64_t val);
+#endif
+
 #if AMU_GROUP1_NR_COUNTERS
 bool amu_group1_supported(void);
 
@@ -86,6 +98,12 @@ bool amu_group1_supported(void);
 uint64_t amu_group1_cnt_read(unsigned int idx);
 void amu_group1_cnt_write(unsigned int idx, uint64_t val);
 void amu_group1_set_evtype(unsigned int idx, unsigned int val);
+
+#if __aarch64__
+uint64_t amu_group1_voffset_read(unsigned int idx);
+void amu_group1_voffset_write(unsigned int idx, uint64_t val);
+#endif
+
 #endif
 
 #endif /* AMU_H */

include/lib/extensions/amu_private.h

 /*
- * Copyright (c) 2017-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2017-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -16,4 +16,12 @@ uint64_t amu_group1_cnt_read_internal(unsigned int idx);
 void amu_group1_cnt_write_internal(unsigned int idx, uint64_t val);
 void amu_group1_set_evtype_internal(unsigned int idx, unsigned int val);
 
+#if __aarch64__
+uint64_t amu_group0_voffset_read_internal(unsigned int idx);
+void amu_group0_voffset_write_internal(unsigned int idx, uint64_t val);
+
+uint64_t amu_group1_voffset_read_internal(unsigned int idx);
+void amu_group1_voffset_write_internal(unsigned int idx, uint64_t val);
+#endif
+
 #endif /* AMU_PRIVATE_H */

lib/el3_runtime/aarch64/context_mgmt.c

 /*
- * Copyright (c) 2013-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2013-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -216,6 +216,16 @@ void cm_setup_context(cpu_context_t *ctx, const entry_point_info_t *ep)
 		scr_el3 |= SCR_EEL2_BIT;
 	}
 
+	/*
+	 * FEAT_AMUv1p1 virtual offset registers are only accessible from EL3
+	 * and EL2, when clear, this bit traps accesses from EL2 so we set it
+	 * to 1 when EL2 is present.
+	 */
+	if (is_armv8_6_feat_amuv1p1_present() &&
+		(el_implemented(2) != EL_IMPL_NONE)) {
+		scr_el3 |= SCR_AMVOFFEN_BIT;
+	}
+
 	/*
 	 * Initialise SCTLR_EL1 to the reset value corresponding to the target
 	 * execution state setting all fields rather than relying of the hw.

lib/extensions/amu/aarch32/amu.c

 /*
- * Copyright (c) 2017-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2017-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -18,13 +18,17 @@
 
 static struct amu_ctx amu_ctxs[PLATFORM_CORE_COUNT];
 
-/* Check if AMUv1 for Armv8.4 or 8.6 is implemented */
-bool amu_supported(void)
+/*
+ * Get AMU version value from pfr0.
+ * Return values
+ *   ID_PFR0_AMU_V1: FEAT_AMUv1 supported (introduced in ARM v8.4)
+ *   ID_PFR0_AMU_V1P1: FEAT_AMUv1p1 supported (introduced in ARM v8.6)
+ *   ID_PFR0_AMU_NOT_SUPPORTED: not supported
+ */
+unsigned int amu_get_version(void)
 {
-	uint32_t features = read_id_pfr0() >> ID_PFR0_AMU_SHIFT;
-
-	features &= ID_PFR0_AMU_MASK;
-	return ((features == 1U) || (features == 2U));
+	return (unsigned int)(read_id_pfr0() >> ID_PFR0_AMU_SHIFT) &
+		ID_PFR0_AMU_MASK;
 }
 
 #if AMU_GROUP1_NR_COUNTERS
@@ -43,7 +47,7 @@ bool amu_group1_supported(void)
  */
 void amu_enable(bool el2_unused)
 {
-	if (!amu_supported()) {
+	if (amu_get_version() == ID_PFR0_AMU_NOT_SUPPORTED) {
 		return;
 	}
 
@@ -87,12 +91,31 @@ void amu_enable(bool el2_unused)
 	/* Enable group 1 counters */
 	write_amcntenset1(AMU_GROUP1_COUNTERS_MASK);
 #endif
+
+	/* Initialize FEAT_AMUv1p1 features if present. */
+	if (amu_get_version() < ID_PFR0_AMU_V1P1) {
+		return;
+	}
+
+#if AMU_RESTRICT_COUNTERS
+	/*
+	 * FEAT_AMUv1p1 adds a register field to restrict access to group 1
+	 * counters at all but the highest implemented EL.  This is controlled
+	 * with the AMU_RESTRICT_COUNTERS compile time flag, when set, system
+	 * register reads at lower ELs return zero.  Reads from the memory
+	 * mapped view are unaffected.
+	 */
+	VERBOSE("AMU group 1 counter access restricted.\n");
+	write_amcr(read_amcr() | AMCR_CG1RZ_BIT);
+#else
+	write_amcr(read_amcr() & ~AMCR_CG1RZ_BIT);
+#endif
 }
 
 /* Read the group 0 counter identified by the given `idx`. */
 uint64_t amu_group0_cnt_read(unsigned int idx)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_PFR0_AMU_NOT_SUPPORTED);
 	assert(idx < AMU_GROUP0_NR_COUNTERS);
 
 	return amu_group0_cnt_read_internal(idx);
@@ -101,7 +124,7 @@ uint64_t amu_group0_cnt_read(unsigned int idx)
 /* Write the group 0 counter identified by the given `idx` with `val` */
 void amu_group0_cnt_write(unsigned  int idx, uint64_t val)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_PFR0_AMU_NOT_SUPPORTED);
 	assert(idx < AMU_GROUP0_NR_COUNTERS);
 
 	amu_group0_cnt_write_internal(idx, val);
@@ -112,7 +135,7 @@ void amu_group0_cnt_write(unsigned  int idx, uint64_t val)
 /* Read the group 1 counter identified by the given `idx` */
 uint64_t amu_group1_cnt_read(unsigned  int idx)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_PFR0_AMU_NOT_SUPPORTED);
 	assert(amu_group1_supported());
 	assert(idx < AMU_GROUP1_NR_COUNTERS);
 
@@ -122,7 +145,7 @@ uint64_t amu_group1_cnt_read(unsigned  int idx)
 /* Write the group 1 counter identified by the given `idx` with `val` */
 void amu_group1_cnt_write(unsigned  int idx, uint64_t val)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_PFR0_AMU_NOT_SUPPORTED);
 	assert(amu_group1_supported());
 	assert(idx < AMU_GROUP1_NR_COUNTERS);
 
@@ -136,7 +159,7 @@ void amu_group1_cnt_write(unsigned  int idx, uint64_t val)
  */
 void amu_group1_set_evtype(unsigned int idx, unsigned int val)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_PFR0_AMU_NOT_SUPPORTED);
 	assert(amu_group1_supported());
 	assert(idx < AMU_GROUP1_NR_COUNTERS);
 
@@ -150,7 +173,7 @@ static void *amu_context_save(const void *arg)
 	struct amu_ctx *ctx = &amu_ctxs[plat_my_core_pos()];
 	unsigned int i;
 
-	if (!amu_supported()) {
+	if (amu_get_version() == ID_PFR0_AMU_NOT_SUPPORTED) {
 		return (void *)-1;
 	}
 
@@ -197,7 +220,7 @@ static void *amu_context_restore(const void *arg)
 	struct amu_ctx *ctx = &amu_ctxs[plat_my_core_pos()];
 	unsigned int i;
 
-	if (!amu_supported()) {
+	if (amu_get_version() == ID_PFR0_AMU_NOT_SUPPORTED) {
 		return (void *)-1;
 	}
 

lib/extensions/amu/aarch32/amu_helpers.S

 /*
- * Copyright (c) 2018, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -75,13 +75,13 @@ func amu_group0_cnt_write_internal
 
 1:
 	stcopr16	r2, r3, AMEVCNTR00	/* index 0 */
-	bx 		lr
+	bx		lr
 	stcopr16	r2, r3, AMEVCNTR01	/* index 1 */
-	bx 		lr
+	bx		lr
 	stcopr16	r2, r3, AMEVCNTR02	/* index 2 */
-	bx 		lr
+	bx		lr
 	stcopr16	r2, r3, AMEVCNTR03	/* index 3 */
-	bx 		lr
+	bx		lr
 endfunc amu_group0_cnt_write_internal
 
 /*
@@ -169,37 +169,37 @@ func amu_group1_cnt_write_internal
 	bx	r1
 
 1:
-	stcopr16	r2, r3,	AMEVCNTR10	/* index 0 */
+	stcopr16	r2, r3, AMEVCNTR10	/* index 0 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR11	/* index 1 */
+	stcopr16	r2, r3, AMEVCNTR11	/* index 1 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR12	/* index 2 */
+	stcopr16	r2, r3, AMEVCNTR12	/* index 2 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR13	/* index 3 */
+	stcopr16	r2, r3, AMEVCNTR13	/* index 3 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR14	/* index 4 */
+	stcopr16	r2, r3, AMEVCNTR14	/* index 4 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR15	/* index 5 */
+	stcopr16	r2, r3, AMEVCNTR15	/* index 5 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR16	/* index 6 */
+	stcopr16	r2, r3, AMEVCNTR16	/* index 6 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR17	/* index 7 */
+	stcopr16	r2, r3, AMEVCNTR17	/* index 7 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR18	/* index 8 */
+	stcopr16	r2, r3, AMEVCNTR18	/* index 8 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR19	/* index 9 */
+	stcopr16	r2, r3, AMEVCNTR19	/* index 9 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR1A	/* index 10 */
+	stcopr16	r2, r3, AMEVCNTR1A	/* index 10 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR1B	/* index 11 */
+	stcopr16	r2, r3, AMEVCNTR1B	/* index 11 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR1C	/* index 12 */
+	stcopr16	r2, r3, AMEVCNTR1C	/* index 12 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR1D	/* index 13 */
+	stcopr16	r2, r3, AMEVCNTR1D	/* index 13 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR1E	/* index 14 */
+	stcopr16	r2, r3, AMEVCNTR1E	/* index 14 */
 	bx		lr
-	stcopr16	r2, r3,	AMEVCNTR1F	/* index 15 */
+	stcopr16	r2, r3, AMEVCNTR1F	/* index 15 */
 	bx		lr
 endfunc amu_group1_cnt_write_internal
 
@@ -234,36 +234,36 @@ func amu_group1_set_evtype_internal
 	bx	r2
 
 1:
-	stcopr	r1,	AMEVTYPER10 /* index 0 */
+	stcopr	r1, AMEVTYPER10 /* index 0 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER11 /* index 1 */
+	stcopr	r1, AMEVTYPER11 /* index 1 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER12 /* index 2 */
+	stcopr	r1, AMEVTYPER12 /* index 2 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER13 /* index 3 */
+	stcopr	r1, AMEVTYPER13 /* index 3 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER14 /* index 4 */
+	stcopr	r1, AMEVTYPER14 /* index 4 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER15 /* index 5 */
+	stcopr	r1, AMEVTYPER15 /* index 5 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER16 /* index 6 */
+	stcopr	r1, AMEVTYPER16 /* index 6 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER17 /* index 7 */
+	stcopr	r1, AMEVTYPER17 /* index 7 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER18 /* index 8 */
+	stcopr	r1, AMEVTYPER18 /* index 8 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER19 /* index 9 */
+	stcopr	r1, AMEVTYPER19 /* index 9 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER1A /* index 10 */
+	stcopr	r1, AMEVTYPER1A /* index 10 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER1B /* index 11 */
+	stcopr	r1, AMEVTYPER1B /* index 11 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER1C /* index 12 */
+	stcopr	r1, AMEVTYPER1C /* index 12 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER1D /* index 13 */
+	stcopr	r1, AMEVTYPER1D /* index 13 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER1E /* index 14 */
+	stcopr	r1, AMEVTYPER1E /* index 14 */
 	bx	lr
-	stcopr	r1,	AMEVTYPER1F /* index 15 */
+	stcopr	r1, AMEVTYPER1F /* index 15 */
 	bx	lr
 endfunc amu_group1_set_evtype_internal

lib/extensions/amu/aarch64/amu.c

 /*
- * Copyright (c) 2017-2020, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2017-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -8,6 +8,7 @@
 #include <stdbool.h>
 
 #include <arch.h>
+#include <arch_features.h>
 #include <arch_helpers.h>
 
 #include <lib/el3_runtime/pubsub_events.h>
@@ -18,13 +19,17 @@
 
 static struct amu_ctx amu_ctxs[PLATFORM_CORE_COUNT];
 
-/* Check if AMUv1 for Armv8.4 or 8.6 is implemented */
-bool amu_supported(void)
+/*
+ * Get AMU version value from aa64pfr0.
+ * Return values
+ *   ID_AA64PFR0_AMU_V1: FEAT_AMUv1 supported (introduced in ARM v8.4)
+ *   ID_AA64PFR0_AMU_V1P1: FEAT_AMUv1p1 supported (introduced in ARM v8.6)
+ *   ID_AA64PFR0_AMU_NOT_SUPPORTED: not supported
+ */
+unsigned int amu_get_version(void)
 {
-	uint64_t features = read_id_aa64pfr0_el1() >> ID_AA64PFR0_AMU_SHIFT;
-
-	features &= ID_AA64PFR0_AMU_MASK;
-	return ((features == 1U) || (features == 2U));
+	return (unsigned int)(read_id_aa64pfr0_el1() >> ID_AA64PFR0_AMU_SHIFT) &
+		ID_AA64PFR0_AMU_MASK;
 }
 
 #if AMU_GROUP1_NR_COUNTERS
@@ -44,8 +49,9 @@ bool amu_group1_supported(void)
 void amu_enable(bool el2_unused)
 {
 	uint64_t v;
+	unsigned int amu_version = amu_get_version();
 
-	if (!amu_supported()) {
+	if (amu_version == ID_AA64PFR0_AMU_NOT_SUPPORTED) {
 		return;
 	}
 
@@ -96,12 +102,36 @@ void amu_enable(bool el2_unused)
 	/* Enable group 1 counters */
 	write_amcntenset1_el0(AMU_GROUP1_COUNTERS_MASK);
 #endif
+
+	/* Initialize FEAT_AMUv1p1 features if present. */
+	if (amu_version < ID_AA64PFR0_AMU_V1P1) {
+		return;
+	}
+
+	if (el2_unused) {
+		/* Make sure virtual offsets are disabled if EL2 not used. */
+		write_hcr_el2(read_hcr_el2() & ~HCR_AMVOFFEN_BIT);
+	}
+
+#if AMU_RESTRICT_COUNTERS
+	/*
+	 * FEAT_AMUv1p1 adds a register field to restrict access to group 1
+	 * counters at all but the highest implemented EL.  This is controlled
+	 * with the AMU_RESTRICT_COUNTERS compile time flag, when set, system
+	 * register reads at lower ELs return zero.  Reads from the memory
+	 * mapped view are unaffected.
+	 */
+	VERBOSE("AMU group 1 counter access restricted.\n");
+	write_amcr_el0(read_amcr_el0() | AMCR_CG1RZ_BIT);
+#else
+	write_amcr_el0(read_amcr_el0() & ~AMCR_CG1RZ_BIT);
+#endif
 }
 
 /* Read the group 0 counter identified by the given `idx`. */
 uint64_t amu_group0_cnt_read(unsigned int idx)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_AA64PFR0_AMU_NOT_SUPPORTED);
 	assert(idx < AMU_GROUP0_NR_COUNTERS);
 
 	return amu_group0_cnt_read_internal(idx);
@@ -110,18 +140,49 @@ uint64_t amu_group0_cnt_read(unsigned int idx)
 /* Write the group 0 counter identified by the given `idx` with `val` */
 void amu_group0_cnt_write(unsigned  int idx, uint64_t val)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_AA64PFR0_AMU_NOT_SUPPORTED);
 	assert(idx < AMU_GROUP0_NR_COUNTERS);
 
 	amu_group0_cnt_write_internal(idx, val);
 	isb();
 }
 
+/*
+ * Read the group 0 offset register for a given index. Index must be 0, 2,
+ * or 3, the register for 1 does not exist.
+ *
+ * Using this function requires FEAT_AMUv1p1 support.
+ */
+uint64_t amu_group0_voffset_read(unsigned int idx)
+{
+	assert(amu_get_version() >= ID_AA64PFR0_AMU_V1P1);
+	assert(idx < AMU_GROUP0_NR_COUNTERS);
+	assert(idx != 1U);
+
+	return amu_group0_voffset_read_internal(idx);
+}
+
+/*
+ * Write the group 0 offset register for a given index. Index must be 0, 2, or
+ * 3, the register for 1 does not exist.
+ *
+ * Using this function requires FEAT_AMUv1p1 support.
+ */
+void amu_group0_voffset_write(unsigned int idx, uint64_t val)
+{
+	assert(amu_get_version() >= ID_AA64PFR0_AMU_V1P1);
+	assert(idx < AMU_GROUP0_NR_COUNTERS);
+	assert(idx != 1U);
+
+	amu_group0_voffset_write_internal(idx, val);
+	isb();
+}
+
 #if AMU_GROUP1_NR_COUNTERS
 /* Read the group 1 counter identified by the given `idx` */
-uint64_t amu_group1_cnt_read(unsigned  int idx)
+uint64_t amu_group1_cnt_read(unsigned int idx)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_AA64PFR0_AMU_NOT_SUPPORTED);
 	assert(amu_group1_supported());
 	assert(idx < AMU_GROUP1_NR_COUNTERS);
 
@@ -129,9 +190,9 @@ uint64_t amu_group1_cnt_read(unsigned  int idx)
 }
 
 /* Write the group 1 counter identified by the given `idx` with `val` */
-void amu_group1_cnt_write(unsigned  int idx, uint64_t val)
+void amu_group1_cnt_write(unsigned int idx, uint64_t val)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_AA64PFR0_AMU_NOT_SUPPORTED);
 	assert(amu_group1_supported());
 	assert(idx < AMU_GROUP1_NR_COUNTERS);
 
@@ -139,13 +200,46 @@ void amu_group1_cnt_write(unsigned  int idx, uint64_t val)
 	isb();
 }
 
+/*
+ * Read the group 1 offset register for a given index.
+ *
+ * Using this function requires FEAT_AMUv1p1 support.
+ */
+uint64_t amu_group1_voffset_read(unsigned int idx)
+{
+	assert(amu_get_version() >= ID_AA64PFR0_AMU_V1P1);
+	assert(amu_group1_supported());
+	assert(idx < AMU_GROUP1_NR_COUNTERS);
+	assert(((read_amcg1idr_el0() >> AMCG1IDR_VOFF_SHIFT) &
+		(1ULL << idx)) != 0ULL);
+
+	return amu_group1_voffset_read_internal(idx);
+}
+
+/*
+ * Write the group 1 offset register for a given index.
+ *
+ * Using this function requires FEAT_AMUv1p1 support.
+ */
+void amu_group1_voffset_write(unsigned int idx, uint64_t val)
+{
+	assert(amu_get_version() >= ID_AA64PFR0_AMU_V1P1);
+	assert(amu_group1_supported());
+	assert(idx < AMU_GROUP1_NR_COUNTERS);
+	assert(((read_amcg1idr_el0() >> AMCG1IDR_VOFF_SHIFT) &
+		(1ULL << idx)) != 0ULL);
+
+	amu_group1_voffset_write_internal(idx, val);
+	isb();
+}
+
 /*
  * Program the event type register for the given `idx` with
  * the event number `val`
  */
 void amu_group1_set_evtype(unsigned int idx, unsigned int val)
 {
-	assert(amu_supported());
+	assert(amu_get_version() != ID_AA64PFR0_AMU_NOT_SUPPORTED);
 	assert(amu_group1_supported());
 	assert(idx < AMU_GROUP1_NR_COUNTERS);
 
@@ -159,7 +253,7 @@ static void *amu_context_save(const void *arg)
 	struct amu_ctx *ctx = &amu_ctxs[plat_my_core_pos()];
 	unsigned int i;
 
-	if (!amu_supported()) {
+	if (amu_get_version() == ID_AA64PFR0_AMU_NOT_SUPPORTED) {
 		return (void *)-1;
 	}
 
@@ -190,13 +284,37 @@ static void *amu_context_save(const void *arg)
 		ctx->group0_cnts[i] = amu_group0_cnt_read(i);
 	}
 
+	/* Save group 0 virtual offsets if supported and enabled. */
+	if ((amu_get_version() >= ID_AA64PFR0_AMU_V1P1) &&
+			((read_hcr_el2() & HCR_AMVOFFEN_BIT) != 0ULL)) {
+		/* Not using a loop because count is fixed and index 1 DNE. */
+		ctx->group0_voffsets[0U] = amu_group0_voffset_read(0U);
+		ctx->group0_voffsets[1U] = amu_group0_voffset_read(2U);
+		ctx->group0_voffsets[2U] = amu_group0_voffset_read(3U);
+	}
+
 #if AMU_GROUP1_NR_COUNTERS
 	/* Save group 1 counters */
 	for (i = 0U; i < AMU_GROUP1_NR_COUNTERS; i++) {
-		if ((AMU_GROUP1_COUNTERS_MASK & (1U << i)) != 0U) {
+		if ((AMU_GROUP1_COUNTERS_MASK & (1UL << i)) != 0U) {
 			ctx->group1_cnts[i] = amu_group1_cnt_read(i);
 		}
 	}
+
+	/* Save group 1 virtual offsets if supported and enabled. */
+	if ((amu_get_version() >= ID_AA64PFR0_AMU_V1P1) &&
+			((read_hcr_el2() & HCR_AMVOFFEN_BIT) != 0ULL)) {
+		u_register_t amcg1idr = read_amcg1idr_el0() >>
+			AMCG1IDR_VOFF_SHIFT;
+		amcg1idr = amcg1idr & AMU_GROUP1_COUNTERS_MASK;
+
+		for (i = 0U; i < AMU_GROUP1_NR_COUNTERS; i++) {
+			if (((amcg1idr >> i) & 1ULL) != 0ULL) {
+				ctx->group1_voffsets[i] =
+					amu_group1_voffset_read(i);
+			}
+		}
+	}
 #endif
 	return (void *)0;
 }
@@ -206,7 +324,7 @@ static void *amu_context_restore(const void *arg)
 	struct amu_ctx *ctx = &amu_ctxs[plat_my_core_pos()];
 	unsigned int i;
 
-	if (!amu_supported()) {
+	if (amu_get_version() == ID_AA64PFR0_AMU_NOT_SUPPORTED) {
 		return (void *)-1;
 	}
 
@@ -227,17 +345,41 @@ static void *amu_context_restore(const void *arg)
 		amu_group0_cnt_write(i, ctx->group0_cnts[i]);
 	}
 
+	/* Restore group 0 virtual offsets if supported and enabled. */
+	if ((amu_get_version() >= ID_AA64PFR0_AMU_V1P1) &&
+			((read_hcr_el2() & HCR_AMVOFFEN_BIT) != 0ULL)) {
+		/* Not using a loop because count is fixed and index 1 DNE. */
+		amu_group0_voffset_write(0U, ctx->group0_voffsets[0U]);
+		amu_group0_voffset_write(2U, ctx->group0_voffsets[1U]);
+		amu_group0_voffset_write(3U, ctx->group0_voffsets[2U]);
+	}
+
 	/* Restore group 0 counter configuration */
 	write_amcntenset0_el0(AMU_GROUP0_COUNTERS_MASK);
 
 #if AMU_GROUP1_NR_COUNTERS
 	/* Restore group 1 counters */
 	for (i = 0U; i < AMU_GROUP1_NR_COUNTERS; i++) {
-		if ((AMU_GROUP1_COUNTERS_MASK & (1U << i)) != 0U) {
+		if ((AMU_GROUP1_COUNTERS_MASK & (1UL << i)) != 0U) {
 			amu_group1_cnt_write(i, ctx->group1_cnts[i]);
 		}
 	}
 
+	/* Restore group 1 virtual offsets if supported and enabled. */
+	if ((amu_get_version() >= ID_AA64PFR0_AMU_V1P1) &&
+			((read_hcr_el2() & HCR_AMVOFFEN_BIT) != 0ULL)) {
+		u_register_t amcg1idr = read_amcg1idr_el0() >>
+			AMCG1IDR_VOFF_SHIFT;
+		amcg1idr = amcg1idr & AMU_GROUP1_COUNTERS_MASK;
+
+		for (i = 0U; i < AMU_GROUP1_NR_COUNTERS; i++) {
+			if (((amcg1idr >> i) & 1ULL) != 0ULL) {
+				amu_group1_voffset_write(i,
+					ctx->group1_voffsets[i]);
+			}
+		}
+	}
+
 	/* Restore group 1 counter configuration */
 	write_amcntenset1_el0(AMU_GROUP1_COUNTERS_MASK);
 #endif

lib/extensions/amu/aarch64/amu_helpers.S

 /*
- * Copyright (c) 2017-2019, ARM Limited and Contributors. All rights reserved.
+ * Copyright (c) 2017-2021, ARM Limited and Contributors. All rights reserved.
  *
  * SPDX-License-Identifier: BSD-3-Clause
  */
@@ -14,6 +14,12 @@
 	.globl	amu_group1_cnt_write_internal
 	.globl	amu_group1_set_evtype_internal
 
+	/* FEAT_AMUv1p1 virtualisation offset register functions */
+	.globl	amu_group0_voffset_read_internal
+	.globl	amu_group0_voffset_write_internal
+	.globl	amu_group1_voffset_read_internal
+	.globl	amu_group1_voffset_write_internal
+
 /*
  * uint64_t amu_group0_cnt_read_internal(int idx);
  *
@@ -211,3 +217,169 @@ func amu_group1_set_evtype_internal
 	write	AMEVTYPER1E_EL0		/* index 14 */
 	write	AMEVTYPER1F_EL0		/* index 15 */
 endfunc amu_group1_set_evtype_internal
+
+/*
+ * Accessor functions for virtual offset registers added with FEAT_AMUv1p1
+ */
+
+/*
+ * uint64_t amu_group0_voffset_read_internal(int idx);
+ *
+ * Given `idx`, read the corresponding AMU virtual offset register
+ * and return it in `x0`.
+ */
+func amu_group0_voffset_read_internal
+	adr	x1, 1f
+#if ENABLE_ASSERTIONS
+	/*
+	 * It can be dangerous to call this function with an
+	 * out of bounds index.  Ensure `idx` is valid.
+	 */
+	tst	x0, #~3
+	ASM_ASSERT(eq)
+	/* Make sure idx != 1 since AMEVCNTVOFF01_EL2 does not exist */
+	cmp	x0, #1
+	ASM_ASSERT(ne)
+#endif
+	/*
+	 * Given `idx` calculate address of mrs/ret instruction pair
+	 * in the table below.
+	 */
+	add	x1, x1, x0, lsl #3	/* each mrs/ret sequence is 8 bytes */
+#if ENABLE_BTI
+	add	x1, x1, x0, lsl #2	/* + "bti j" instruction */
+#endif
+	br	x1
+
+1:	read	AMEVCNTVOFF00_EL2	/* index 0 */
+	.skip	8			/* AMEVCNTVOFF01_EL2 does not exist */
+#if ENABLE_BTI
+	.skip	4
+#endif
+	read	AMEVCNTVOFF02_EL2	/* index 2 */
+	read	AMEVCNTVOFF03_EL2	/* index 3 */
+endfunc amu_group0_voffset_read_internal
+
+/*
+ * void amu_group0_voffset_write_internal(int idx, uint64_t val);
+ *
+ * Given `idx`, write `val` to the corresponding AMU virtual offset register.
+ */
+func amu_group0_voffset_write_internal
+	adr	x2, 1f
+#if ENABLE_ASSERTIONS
+	/*
+	 * It can be dangerous to call this function with an
+	 * out of bounds index.  Ensure `idx` is valid.
+	 */
+	tst	x0, #~3
+	ASM_ASSERT(eq)
+	/* Make sure idx != 1 since AMEVCNTVOFF01_EL2 does not exist */
+	cmp	x0, #1
+	ASM_ASSERT(ne)
+#endif
+	/*
+	 * Given `idx` calculate address of mrs/ret instruction pair
+	 * in the table below.
+	 */
+	add	x2, x2, x0, lsl #3	/* each msr/ret sequence is 8 bytes */
+#if ENABLE_BTI
+	add	x2, x2, x0, lsl #2	/* + "bti j" instruction */
+#endif
+	br	x2
+
+1:	write	AMEVCNTVOFF00_EL2	/* index 0 */
+	.skip	8			/* AMEVCNTVOFF01_EL2 does not exist */
+#if ENABLE_BTI
+	.skip	4
+#endif
+	write	AMEVCNTVOFF02_EL2	/* index 2 */
+	write	AMEVCNTVOFF03_EL2	/* index 3 */
+endfunc amu_group0_voffset_write_internal
+
+/*
+ * uint64_t amu_group1_voffset_read_internal(int idx);
+ *
+ * Given `idx`, read the corresponding AMU virtual offset register
+ * and return it in `x0`.
+ */
+func amu_group1_voffset_read_internal
+	adr	x1, 1f
+#if ENABLE_ASSERTIONS
+	/*
+	 * It can be dangerous to call this function with an
+	 * out of bounds index.  Ensure `idx` is valid.
+	 */
+	tst	x0, #~0xF
+	ASM_ASSERT(eq)
+#endif
+	/*
+	 * Given `idx` calculate address of mrs/ret instruction pair
+	 * in the table below.
+	 */
+	add	x1, x1, x0, lsl #3	/* each mrs/ret sequence is 8 bytes */
+#if ENABLE_BTI
+	add	x1, x1, x0, lsl #2	/* + "bti j" instruction */
+#endif
+	br	x1
+
+1:	read	AMEVCNTVOFF10_EL2	/* index 0 */
+	read	AMEVCNTVOFF11_EL2	/* index 1 */
+	read	AMEVCNTVOFF12_EL2	/* index 2 */
+	read	AMEVCNTVOFF13_EL2	/* index 3 */
+	read	AMEVCNTVOFF14_EL2	/* index 4 */
+	read	AMEVCNTVOFF15_EL2	/* index 5 */
+	read	AMEVCNTVOFF16_EL2	/* index 6 */
+	read	AMEVCNTVOFF17_EL2	/* index 7 */
+	read	AMEVCNTVOFF18_EL2	/* index 8 */
+	read	AMEVCNTVOFF19_EL2	/* index 9 */
+	read	AMEVCNTVOFF1A_EL2	/* index 10 */
+	read	AMEVCNTVOFF1B_EL2	/* index 11 */
+	read	AMEVCNTVOFF1C_EL2	/* index 12 */
+	read	AMEVCNTVOFF1D_EL2	/* index 13 */
+	read	AMEVCNTVOFF1E_EL2	/* index 14 */
+	read	AMEVCNTVOFF1F_EL2	/* index 15 */
+endfunc amu_group1_voffset_read_internal
+
+/*
+ * void amu_group1_voffset_write_internal(int idx, uint64_t val);
+ *
+ * Given `idx`, write `val` to the corresponding AMU virtual offset register.
+ */
+func amu_group1_voffset_write_internal
+	adr	x2, 1f
+#if ENABLE_ASSERTIONS
+	/*
+	 * It can be dangerous to call this function with an
+	 * out of bounds index.  Ensure `idx` is valid.
+	 */
+	tst	x0, #~0xF
+	ASM_ASSERT(eq)
+#endif
+	/*
+	 * Given `idx` calculate address of mrs/ret instruction pair
+	 * in the table below.
+	 */
+	add	x2, x2, x0, lsl #3	/* each msr/ret sequence is 8 bytes */
+#if ENABLE_BTI
+	add	x2, x2, x0, lsl #2	/* + "bti j" instruction */
+#endif
+	br	x2
+
+1:	write	AMEVCNTVOFF10_EL2	/* index 0 */
+	write	AMEVCNTVOFF11_EL2	/* index 1 */
+	write	AMEVCNTVOFF12_EL2	/* index 2 */
+	write	AMEVCNTVOFF13_EL2	/* index 3 */
+	write	AMEVCNTVOFF14_EL2	/* index 4 */
+	write	AMEVCNTVOFF15_EL2	/* index 5 */
+	write	AMEVCNTVOFF16_EL2	/* index 6 */
+	write	AMEVCNTVOFF17_EL2	/* index 7 */
+	write	AMEVCNTVOFF18_EL2	/* index 8 */
+	write	AMEVCNTVOFF19_EL2	/* index 9 */
+	write	AMEVCNTVOFF1A_EL2	/* index 10 */
+	write	AMEVCNTVOFF1B_EL2	/* index 11 */
+	write	AMEVCNTVOFF1C_EL2	/* index 12 */
+	write	AMEVCNTVOFF1D_EL2	/* index 13 */
+	write	AMEVCNTVOFF1E_EL2	/* index 14 */
+	write	AMEVCNTVOFF1F_EL2	/* index 15 */
+endfunc amu_group1_voffset_write_internal

make_helpers/defaults.mk

@@ -291,9 +291,10 @@ endif
 # Include Memory Tagging Extension registers in cpu context. This must be set
 # to 1 if the platform wants to use this feature in the Secure world and MTE is
 # enabled at ELX.
-CTX_INCLUDE_MTE_REGS := 0
+CTX_INCLUDE_MTE_REGS		:= 0
 
 ENABLE_AMU			:= 0
+AMU_RESTRICT_COUNTERS		:= 0
 
 # By default, enable Scalable Vector Extension if implemented for Non-secure
 # lower ELs