Merge tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux

Pull arm64 updates from Will Deacon:

 - RAS reporting via GHES/APEI (ACPI)

 - Indirect ftrace trampolines for modules

 - Improvements to kernel fault reporting

 - Page poisoning

 - Sigframe cleanups and preparation for SVE context

 - Core dump fixes

 - Sparse fixes (mainly relating to endianness)

 - xgene SoC PMU v3 driver

 - Misc cleanups and non-critical fixes

* tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux: (75 commits)
  arm64: fix endianness annotation for 'struct jit_ctx' and friends
  arm64: cpuinfo: constify attribute_group structures.
  arm64: ptrace: Fix incorrect get_user() use in compat_vfp_set()
  arm64: ptrace: Remove redundant overrun check from compat_vfp_set()
  arm64: ptrace: Avoid setting compat FP[SC]R to garbage if get_user fails
  arm64: fix endianness annotation for __apply_alternatives()/get_alt_insn()
  arm64: fix endianness annotation in get_kaslr_seed()
  arm64: add missing conversion to __wsum in ip_fast_csum()
  arm64: fix endianness annotation in acpi_parking_protocol.c
  arm64: use readq() instead of readl() to read 64bit entry_point
  arm64: fix endianness annotation for reloc_insn_movw() & reloc_insn_imm()
  arm64: fix endianness annotation for aarch64_insn_write()
  arm64: fix endianness annotation in aarch64_insn_read()
  arm64: fix endianness annotation in call_undef_hook()
  arm64: fix endianness annotation for debug-monitors.c
  ras: mark stub functions as 'inline'
  arm64: pass endianness info to sparse
  arm64: ftrace: fix !CONFIG_ARM64_MODULE_PLTS kernels
  arm64: signal: Allow expansion of the signal frame
  acpi: apei: check for pending errors when probing GHES entries
  ...

10 files changed, 1033 insertions, 126 deletions

diff --git a/drivers/acpi/apei/Kconfig b/drivers/acpi/apei/Kconfig
index b0140c8fc733..de14d49a5c90 100644
--- a/drivers/acpi/apei/Kconfig
+++ b/drivers/acpi/apei/Kconfig
@@ -39,6 +39,21 @@ config ACPI_APEI_PCIEAER
 	  PCIe AER errors may be reported via APEI firmware first mode.
 	  Turn on this option to enable the corresponding support.
 
+config ACPI_APEI_SEA
+	bool "APEI Synchronous External Abort logging/recovering support"
+	depends on ARM64 && ACPI_APEI_GHES
+	default y
+	help
+	  This option should be enabled if the system supports
+	  firmware first handling of SEA (Synchronous External Abort).
+	  SEA happens with certain faults of data abort or instruction
+	  abort synchronous exceptions on ARMv8 systems. If a system
+	  supports firmware first handling of SEA, the platform analyzes
+	  and handles hardware error notifications from SEA, and it may then
+	  form a HW error record for the OS to parse and handle. This
+	  option allows the OS to look for such hardware error record, and
+	  take appropriate action.
+
 config ACPI_APEI_MEMORY_FAILURE
 	bool "APEI memory error recovering support"
 	depends on ACPI_APEI && MEMORY_FAILURE
diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index 0968816f5755..d661d452b238 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -45,10 +45,14 @@
 #include <linux/aer.h>
 #include <linux/nmi.h>
 #include <linux/sched/clock.h>
+#include <linux/uuid.h>
+#include <linux/ras.h>
 
+#include <acpi/actbl1.h>
 #include <acpi/ghes.h>
 #include <acpi/apei.h>
 #include <asm/tlbflush.h>
+#include <ras/ras_event.h>
 
 #include "apei-internal.h"
 
@@ -80,6 +84,11 @@
 	((struct acpi_hest_generic_status *)				\
 	 ((struct ghes_estatus_node *)(estatus_node) + 1))
 
+static inline bool is_hest_type_generic_v2(struct ghes *ghes)
+{
+	return ghes->generic->header.type == ACPI_HEST_TYPE_GENERIC_ERROR_V2;
+}
+
 /*
  * This driver isn't really modular, however for the time being,
  * continuing to use module_param is the easiest way to remain
@@ -110,11 +119,7 @@ static DEFINE_MUTEX(ghes_list_mutex);
  * Two virtual pages are used, one for IRQ/PROCESS context, the other for
  * NMI context (optionally).
  */
-#ifdef CONFIG_HAVE_ACPI_APEI_NMI
 #define GHES_IOREMAP_PAGES           2
-#else
-#define GHES_IOREMAP_PAGES           1
-#endif
 #define GHES_IOREMAP_IRQ_PAGE(base)	(base)
 #define GHES_IOREMAP_NMI_PAGE(base)	((base) + PAGE_SIZE)
 
@@ -133,6 +138,8 @@ static unsigned long ghes_estatus_pool_size_request;
 static struct ghes_estatus_cache *ghes_estatus_caches[GHES_ESTATUS_CACHES_SIZE];
 static atomic_t ghes_estatus_cache_alloced;
 
+static int ghes_panic_timeout __read_mostly = 30;
+
 static int ghes_ioremap_init(void)
 {
 	ghes_ioremap_area = __get_vm_area(PAGE_SIZE * GHES_IOREMAP_PAGES,
@@ -153,10 +160,14 @@ static void ghes_ioremap_exit(void)
 static void __iomem *ghes_ioremap_pfn_nmi(u64 pfn)
 {
 	unsigned long vaddr;
+	phys_addr_t paddr;
+	pgprot_t prot;
 
 	vaddr = (unsigned long)GHES_IOREMAP_NMI_PAGE(ghes_ioremap_area->addr);
-	ioremap_page_range(vaddr, vaddr + PAGE_SIZE,
-			   pfn << PAGE_SHIFT, PAGE_KERNEL);
+
+	paddr = pfn << PAGE_SHIFT;
+	prot = arch_apei_get_mem_attribute(paddr);
+	ioremap_page_range(vaddr, vaddr + PAGE_SIZE, paddr, prot);
 
 	return (void __iomem *)vaddr;
 }
@@ -240,6 +251,16 @@ static int ghes_estatus_pool_expand(unsigned long len)
 	return 0;
 }
 
+static int map_gen_v2(struct ghes *ghes)
+{
+	return apei_map_generic_address(&ghes->generic_v2->read_ack_register);
+}
+
+static void unmap_gen_v2(struct ghes *ghes)
+{
+	apei_unmap_generic_address(&ghes->generic_v2->read_ack_register);
+}
+
 static struct ghes *ghes_new(struct acpi_hest_generic *generic)
 {
 	struct ghes *ghes;
@@ -249,10 +270,17 @@ static struct ghes *ghes_new(struct acpi_hest_generic *generic)
 	ghes = kzalloc(sizeof(*ghes), GFP_KERNEL);
 	if (!ghes)
 		return ERR_PTR(-ENOMEM);
+
 	ghes->generic = generic;
+	if (is_hest_type_generic_v2(ghes)) {
+		rc = map_gen_v2(ghes);
+		if (rc)
+			goto err_free;
+	}
+
 	rc = apei_map_generic_address(&generic->error_status_address);
 	if (rc)
-		goto err_free;
+		goto err_unmap_read_ack_addr;
 	error_block_length = generic->error_block_length;
 	if (error_block_length > GHES_ESTATUS_MAX_SIZE) {
 		pr_warning(FW_WARN GHES_PFX
@@ -264,13 +292,16 @@ static struct ghes *ghes_new(struct acpi_hest_generic *generic)
 	ghes->estatus = kmalloc(error_block_length, GFP_KERNEL);
 	if (!ghes->estatus) {
 		rc = -ENOMEM;
-		goto err_unmap;
+		goto err_unmap_status_addr;
 	}
 
 	return ghes;
 
-err_unmap:
+err_unmap_status_addr:
 	apei_unmap_generic_address(&generic->error_status_address);
+err_unmap_read_ack_addr:
+	if (is_hest_type_generic_v2(ghes))
+		unmap_gen_v2(ghes);
 err_free:
 	kfree(ghes);
 	return ERR_PTR(rc);
@@ -280,6 +311,8 @@ static void ghes_fini(struct ghes *ghes)
 {
 	kfree(ghes->estatus);
 	apei_unmap_generic_address(&ghes->generic->error_status_address);
+	if (is_hest_type_generic_v2(ghes))
+		unmap_gen_v2(ghes);
 }
 
 static inline int ghes_severity(int severity)
@@ -400,8 +433,7 @@ static void ghes_handle_memory_failure(struct acpi_hest_generic_data *gdata, int
 	unsigned long pfn;
 	int flags = -1;
 	int sec_sev = ghes_severity(gdata->error_severity);
-	struct cper_sec_mem_err *mem_err;
-	mem_err = (struct cper_sec_mem_err *)(gdata + 1);
+	struct cper_sec_mem_err *mem_err = acpi_hest_get_payload(gdata);
 
 	if (!(mem_err->validation_bits & CPER_MEM_VALID_PA))
 		return;
@@ -432,14 +464,22 @@ static void ghes_do_proc(struct ghes *ghes,
 	int sev, sec_sev;
 	struct acpi_hest_generic_data *gdata;
 	guid_t *sec_type;
+	guid_t *fru_id = &NULL_UUID_LE;
+	char *fru_text = "";
 
 	sev = ghes_severity(estatus->error_severity);
 	apei_estatus_for_each_section(estatus, gdata) {
 		sec_type = (guid_t *)gdata->section_type;
 		sec_sev = ghes_severity(gdata->error_severity);
+		if (gdata->validation_bits & CPER_SEC_VALID_FRU_ID)
+			fru_id = (guid_t *)gdata->fru_id;
+
+		if (gdata->validation_bits & CPER_SEC_VALID_FRU_TEXT)
+			fru_text = gdata->fru_text;
+
 		if (guid_equal(sec_type, &CPER_SEC_PLATFORM_MEM)) {
-			struct cper_sec_mem_err *mem_err;
-			mem_err = (struct cper_sec_mem_err *)(gdata+1);
+			struct cper_sec_mem_err *mem_err = acpi_hest_get_payload(gdata);
+
 			ghes_edac_report_mem_error(ghes, sev, mem_err);
 
 			arch_apei_report_mem_error(sev, mem_err);
@@ -447,8 +487,8 @@ static void ghes_do_proc(struct ghes *ghes,
 		}
 #ifdef CONFIG_ACPI_APEI_PCIEAER
 		else if (guid_equal(sec_type, &CPER_SEC_PCIE)) {
-			struct cper_sec_pcie *pcie_err;
-			pcie_err = (struct cper_sec_pcie *)(gdata+1);
+			struct cper_sec_pcie *pcie_err = acpi_hest_get_payload(gdata);
+
 			if (sev == GHES_SEV_RECOVERABLE &&
 			    sec_sev == GHES_SEV_RECOVERABLE &&
 			    pcie_err->validation_bits & CPER_PCIE_VALID_DEVICE_ID &&
@@ -477,6 +517,17 @@ static void ghes_do_proc(struct ghes *ghes,
 
 		}
 #endif
+		else if (guid_equal(sec_type, &CPER_SEC_PROC_ARM)) {
+			struct cper_sec_proc_arm *err = acpi_hest_get_payload(gdata);
+
+			log_arm_hw_error(err);
+		} else {
+			void *err = acpi_hest_get_payload(gdata);
+
+			log_non_standard_event(sec_type, fru_id, fru_text,
+					       sec_sev, err,
+					       gdata->error_data_length);
+		}
 	}
 }
 
@@ -649,6 +700,31 @@ static void ghes_estatus_cache_add(
 	rcu_read_unlock();
 }
 
+static int ghes_ack_error(struct acpi_hest_generic_v2 *gv2)
+{
+	int rc;
+	u64 val = 0;
+
+	rc = apei_read(&val, &gv2->read_ack_register);
+	if (rc)
+		return rc;
+
+	val &= gv2->read_ack_preserve << gv2->read_ack_register.bit_offset;
+	val |= gv2->read_ack_write    << gv2->read_ack_register.bit_offset;
+
+	return apei_write(val, &gv2->read_ack_register);
+}
+
+static void __ghes_panic(struct ghes *ghes)
+{
+	__ghes_print_estatus(KERN_EMERG, ghes->generic, ghes->estatus);
+
+	/* reboot to log the error! */
+	if (!panic_timeout)
+		panic_timeout = ghes_panic_timeout;
+	panic("Fatal hardware error!");
+}
+
 static int ghes_proc(struct ghes *ghes)
 {
 	int rc;
@@ -656,11 +732,26 @@ static int ghes_proc(struct ghes *ghes)
 	rc = ghes_read_estatus(ghes, 0);
 	if (rc)
 		goto out;
+
+	if (ghes_severity(ghes->estatus->error_severity) >= GHES_SEV_PANIC) {
+		__ghes_panic(ghes);
+	}
+
 	if (!ghes_estatus_cached(ghes->estatus)) {
 		if (ghes_print_estatus(NULL, ghes->generic, ghes->estatus))
 			ghes_estatus_cache_add(ghes->generic, ghes->estatus);
 	}
 	ghes_do_proc(ghes, ghes->estatus);
+
+	/*
+	 * GHESv2 type HEST entries introduce support for error acknowledgment,
+	 * so only acknowledge the error if this support is present.
+	 */
+	if (is_hest_type_generic_v2(ghes)) {
+		rc = ghes_ack_error(ghes->generic_v2);
+		if (rc)
+			return rc;
+	}
 out:
 	ghes_clear_estatus(ghes);
 	return rc;
@@ -722,6 +813,55 @@ static struct notifier_block ghes_notifier_hed = {
 	.notifier_call = ghes_notify_hed,
 };
 
+#ifdef CONFIG_ACPI_APEI_SEA
+static LIST_HEAD(ghes_sea);
+
+/*
+ * Return 0 only if one of the SEA error sources successfully reported an error
+ * record sent from the firmware.
+ */
+int ghes_notify_sea(void)
+{
+	struct ghes *ghes;
+	int ret = -ENOENT;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(ghes, &ghes_sea, list) {
+		if (!ghes_proc(ghes))
+			ret = 0;
+	}
+	rcu_read_unlock();
+	return ret;
+}
+
+static void ghes_sea_add(struct ghes *ghes)
+{
+	mutex_lock(&ghes_list_mutex);
+	list_add_rcu(&ghes->list, &ghes_sea);
+	mutex_unlock(&ghes_list_mutex);
+}
+
+static void ghes_sea_remove(struct ghes *ghes)
+{
+	mutex_lock(&ghes_list_mutex);
+	list_del_rcu(&ghes->list);
+	mutex_unlock(&ghes_list_mutex);
+	synchronize_rcu();
+}
+#else /* CONFIG_ACPI_APEI_SEA */
+static inline void ghes_sea_add(struct ghes *ghes)
+{
+	pr_err(GHES_PFX "ID: %d, trying to add SEA notification which is not supported\n",
+	       ghes->generic->header.source_id);
+}
+
+static inline void ghes_sea_remove(struct ghes *ghes)
+{
+	pr_err(GHES_PFX "ID: %d, trying to remove SEA notification which is not supported\n",
+	       ghes->generic->header.source_id);
+}
+#endif /* CONFIG_ACPI_APEI_SEA */
+
 #ifdef CONFIG_HAVE_ACPI_APEI_NMI
 /*
  * printk is not safe in NMI context.  So in NMI handler, we allocate
@@ -742,8 +882,6 @@ static atomic_t ghes_in_nmi = ATOMIC_INIT(0);
 
 static LIST_HEAD(ghes_nmi);
 
-static int ghes_panic_timeout	__read_mostly = 30;
-
 static void ghes_proc_in_irq(struct irq_work *irq_work)
 {
 	struct llist_node *llnode, *next;
@@ -829,18 +967,6 @@ static void __process_error(struct ghes *ghes)
 #endif
 }
 
-static void __ghes_panic(struct ghes *ghes)
-{
-	oops_begin();
-	ghes_print_queued_estatus();
-	__ghes_print_estatus(KERN_EMERG, ghes->generic, ghes->estatus);
-
-	/* reboot to log the error! */
-	if (panic_timeout == 0)
-		panic_timeout = ghes_panic_timeout;
-	panic("Fatal hardware error!");
-}
-
 static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs)
 {
 	struct ghes *ghes;
@@ -858,8 +984,11 @@ static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs)
 		}
 
 		sev = ghes_severity(ghes->estatus->error_severity);
-		if (sev >= GHES_SEV_PANIC)
+		if (sev >= GHES_SEV_PANIC) {
+			oops_begin();
+			ghes_print_queued_estatus();
 			__ghes_panic(ghes);
+		}
 
 		if (!(ghes->flags & GHES_TO_CLEAR))
 			continue;
@@ -970,6 +1099,14 @@ static int ghes_probe(struct platform_device *ghes_dev)
 	case ACPI_HEST_NOTIFY_GPIO:
 		break;
 
+	case ACPI_HEST_NOTIFY_SEA:
+		if (!IS_ENABLED(CONFIG_ACPI_APEI_SEA)) {
+			pr_warn(GHES_PFX "Generic hardware error source: %d notified via SEA is not supported\n",
+				generic->header.source_id);
+			rc = -ENOTSUPP;
+			goto err;
+		}
+		break;
 	case ACPI_HEST_NOTIFY_NMI:
 		if (!IS_ENABLED(CONFIG_HAVE_ACPI_APEI_NMI)) {
 			pr_warn(GHES_PFX "Generic hardware error source: %d notified via NMI interrupt is not supported!\n",
@@ -1038,6 +1175,9 @@ static int ghes_probe(struct platform_device *ghes_dev)
 		mutex_unlock(&ghes_list_mutex);
 		break;
 
+	case ACPI_HEST_NOTIFY_SEA:
+		ghes_sea_add(ghes);
+		break;
 	case ACPI_HEST_NOTIFY_NMI:
 		ghes_nmi_add(ghes);
 		break;
@@ -1046,6 +1186,9 @@ static int ghes_probe(struct platform_device *ghes_dev)
 	}
 	platform_set_drvdata(ghes_dev, ghes);
 
+	/* Handle any pending errors right away */
+	ghes_proc(ghes);
+
 	return 0;
 err_edac_unreg:
 	ghes_edac_unregister(ghes);
@@ -1085,6 +1228,9 @@ static int ghes_remove(struct platform_device *ghes_dev)
 		synchronize_rcu();
 		break;
 
+	case ACPI_HEST_NOTIFY_SEA:
+		ghes_sea_remove(ghes);
+		break;
 	case ACPI_HEST_NOTIFY_NMI:
 		ghes_nmi_remove(ghes);
 		break;
diff --git a/drivers/acpi/apei/hest.c b/drivers/acpi/apei/hest.c
index 8f2a98e23bba..456b488eb1df 100644
--- a/drivers/acpi/apei/hest.c
+++ b/drivers/acpi/apei/hest.c
@@ -52,6 +52,7 @@ static const int hest_esrc_len_tab[ACPI_HEST_TYPE_RESERVED] = {
 	[ACPI_HEST_TYPE_AER_ENDPOINT] = sizeof(struct acpi_hest_aer),
 	[ACPI_HEST_TYPE_AER_BRIDGE] = sizeof(struct acpi_hest_aer_bridge),
 	[ACPI_HEST_TYPE_GENERIC_ERROR] = sizeof(struct acpi_hest_generic),
+	[ACPI_HEST_TYPE_GENERIC_ERROR_V2] = sizeof(struct acpi_hest_generic_v2),
 };
 
 static int hest_esrc_len(struct acpi_hest_header *hest_hdr)
@@ -141,7 +142,8 @@ static int __init hest_parse_ghes_count(struct acpi_hest_header *hest_hdr, void
 {
 	int *count = data;
 
-	if (hest_hdr->type == ACPI_HEST_TYPE_GENERIC_ERROR)
+	if (hest_hdr->type == ACPI_HEST_TYPE_GENERIC_ERROR ||
+	    hest_hdr->type == ACPI_HEST_TYPE_GENERIC_ERROR_V2)
 		(*count)++;
 	return 0;
 }
@@ -152,7 +154,8 @@ static int __init hest_parse_ghes(struct acpi_hest_header *hest_hdr, void *data)
 	struct ghes_arr *ghes_arr = data;
 	int rc, i;
 
-	if (hest_hdr->type != ACPI_HEST_TYPE_GENERIC_ERROR)
+	if (hest_hdr->type != ACPI_HEST_TYPE_GENERIC_ERROR &&
+	    hest_hdr->type != ACPI_HEST_TYPE_GENERIC_ERROR_V2)
 		return 0;
 
 	if (!((struct acpi_hest_generic *)hest_hdr)->enabled)
diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c
index 797b28dc7b34..d048f72c23f8 100644
--- a/drivers/acpi/arm64/iort.c
+++ b/drivers/acpi/arm64/iort.c
@@ -234,21 +234,6 @@ static struct acpi_iort_node *iort_scan_node(enum acpi_iort_node_type type,
 	return NULL;
 }
 
-static acpi_status
-iort_match_type_callback(struct acpi_iort_node *node, void *context)
-{
-	return AE_OK;
-}
-
-bool iort_node_match(u8 type)
-{
-	struct acpi_iort_node *node;
-
-	node = iort_scan_node(type, iort_match_type_callback, NULL);
-
-	return node != NULL;
-}
-
 static acpi_status iort_match_node_callback(struct acpi_iort_node *node,
 					    void *context)
 {
diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig
index 2af70014ee5a..ccd239ab879f 100644
--- a/drivers/char/Kconfig
+++ b/drivers/char/Kconfig
@@ -17,6 +17,8 @@ config DEVMEM
 
 config DEVKMEM
 	bool "/dev/kmem virtual device support"
+	# On arm64, VMALLOC_START < PAGE_OFFSET, which confuses kmem read/write
+	depends on !ARM64
 	help
 	  Say Y here if you want to support the /dev/kmem device. The
 	  /dev/kmem device is rarely used, but can be used for certain
diff --git a/drivers/firmware/efi/cper.c b/drivers/firmware/efi/cper.c
index d42537425438..48a8f69da42a 100644
--- a/drivers/firmware/efi/cper.c
+++ b/drivers/firmware/efi/cper.c
@@ -32,6 +32,10 @@
 #include <linux/acpi.h>
 #include <linux/pci.h>
 #include <linux/aer.h>
+#include <linux/printk.h>
+#include <linux/bcd.h>
+#include <acpi/ghes.h>
+#include <ras/ras_event.h>
 
 #define INDENT_SP	" "
 
@@ -107,12 +111,15 @@ void cper_print_bits(const char *pfx, unsigned int bits,
 static const char * const proc_type_strs[] = {
 	"IA32/X64",
 	"IA64",
+	"ARM",
 };
 
 static const char * const proc_isa_strs[] = {
 	"IA32",
 	"IA64",
 	"X64",
+	"ARM A32/T32",
+	"ARM A64",
 };
 
 static const char * const proc_error_type_strs[] = {
@@ -181,6 +188,122 @@ static void cper_print_proc_generic(const char *pfx,
 		printk("%s""IP: 0x%016llx\n", pfx, proc->ip);
 }
 
+#if defined(CONFIG_ARM64) || defined(CONFIG_ARM)
+static const char * const arm_reg_ctx_strs[] = {
+	"AArch32 general purpose registers",
+	"AArch32 EL1 context registers",
+	"AArch32 EL2 context registers",
+	"AArch32 secure context registers",
+	"AArch64 general purpose registers",
+	"AArch64 EL1 context registers",
+	"AArch64 EL2 context registers",
+	"AArch64 EL3 context registers",
+	"Misc. system register structure",
+};
+
+static void cper_print_proc_arm(const char *pfx,
+				const struct cper_sec_proc_arm *proc)
+{
+	int i, len, max_ctx_type;
+	struct cper_arm_err_info *err_info;
+	struct cper_arm_ctx_info *ctx_info;
+	char newpfx[64];
+
+	printk("%sMIDR: 0x%016llx\n", pfx, proc->midr);
+
+	len = proc->section_length - (sizeof(*proc) +
+		proc->err_info_num * (sizeof(*err_info)));
+	if (len < 0) {
+		printk("%ssection length: %d\n", pfx, proc->section_length);
+		printk("%ssection length is too small\n", pfx);
+		printk("%sfirmware-generated error record is incorrect\n", pfx);
+		printk("%sERR_INFO_NUM is %d\n", pfx, proc->err_info_num);
+		return;
+	}
+
+	if (proc->validation_bits & CPER_ARM_VALID_MPIDR)
+		printk("%sMultiprocessor Affinity Register (MPIDR): 0x%016llx\n",
+			pfx, proc->mpidr);
+
+	if (proc->validation_bits & CPER_ARM_VALID_AFFINITY_LEVEL)
+		printk("%serror affinity level: %d\n", pfx,
+			proc->affinity_level);
+
+	if (proc->validation_bits & CPER_ARM_VALID_RUNNING_STATE) {
+		printk("%srunning state: 0x%x\n", pfx, proc->running_state);
+		printk("%sPower State Coordination Interface state: %d\n",
+			pfx, proc->psci_state);
+	}
+
+	snprintf(newpfx, sizeof(newpfx), "%s%s", pfx, INDENT_SP);
+
+	err_info = (struct cper_arm_err_info *)(proc + 1);
+	for (i = 0; i < proc->err_info_num; i++) {
+		printk("%sError info structure %d:\n", pfx, i);
+
+		printk("%snum errors: %d\n", pfx, err_info->multiple_error + 1);
+
+		if (err_info->validation_bits & CPER_ARM_INFO_VALID_FLAGS) {
+			if (err_info->flags & CPER_ARM_INFO_FLAGS_FIRST)
+				printk("%sfirst error captured\n", newpfx);
+			if (err_info->flags & CPER_ARM_INFO_FLAGS_LAST)
+				printk("%slast error captured\n", newpfx);
+			if (err_info->flags & CPER_ARM_INFO_FLAGS_PROPAGATED)
+				printk("%spropagated error captured\n",
+				       newpfx);
+			if (err_info->flags & CPER_ARM_INFO_FLAGS_OVERFLOW)
+				printk("%soverflow occurred, error info is incomplete\n",
+				       newpfx);
+		}
+
+		printk("%serror_type: %d, %s\n", newpfx, err_info->type,
+			err_info->type < ARRAY_SIZE(proc_error_type_strs) ?
+			proc_error_type_strs[err_info->type] : "unknown");
+		if (err_info->validation_bits & CPER_ARM_INFO_VALID_ERR_INFO)
+			printk("%serror_info: 0x%016llx\n", newpfx,
+			       err_info->error_info);
+		if (err_info->validation_bits & CPER_ARM_INFO_VALID_VIRT_ADDR)
+			printk("%svirtual fault address: 0x%016llx\n",
+				newpfx, err_info->virt_fault_addr);
+		if (err_info->validation_bits & CPER_ARM_INFO_VALID_PHYSICAL_ADDR)
+			printk("%sphysical fault address: 0x%016llx\n",
+				newpfx, err_info->physical_fault_addr);
+		err_info += 1;
+	}
+
+	ctx_info = (struct cper_arm_ctx_info *)err_info;
+	max_ctx_type = ARRAY_SIZE(arm_reg_ctx_strs) - 1;
+	for (i = 0; i < proc->context_info_num; i++) {
+		int size = sizeof(*ctx_info) + ctx_info->size;
+
+		printk("%sContext info structure %d:\n", pfx, i);
+		if (len < size) {
+			printk("%ssection length is too small\n", newpfx);
+			printk("%sfirmware-generated error record is incorrect\n", pfx);
+			return;
+		}
+		if (ctx_info->type > max_ctx_type) {
+			printk("%sInvalid context type: %d (max: %d)\n",
+				newpfx, ctx_info->type, max_ctx_type);
+			return;
+		}
+		printk("%sregister context type: %s\n", newpfx,
+			arm_reg_ctx_strs[ctx_info->type]);
+		print_hex_dump(newpfx, "", DUMP_PREFIX_OFFSET, 16, 4,
+				(ctx_info + 1), ctx_info->size, 0);
+		len -= size;
+		ctx_info = (struct cper_arm_ctx_info *)((long)ctx_info + size);
+	}
+
+	if (len > 0) {
+		printk("%sVendor specific error info has %u bytes:\n", pfx,
+		       len);
+		print_hex_dump(newpfx, "", DUMP_PREFIX_OFFSET, 16, 4, ctx_info,
+				len, true);
+	}
+}
+#endif
+
 static const char * const mem_err_type_strs[] = {
 	"unknown",
 	"no error",
@@ -386,13 +509,38 @@ static void cper_print_pcie(const char *pfx, const struct cper_sec_pcie *pcie,
 	pfx, pcie->bridge.secondary_status, pcie->bridge.control);
 }
 
-static void cper_estatus_print_section(
-	const char *pfx, const struct acpi_hest_generic_data *gdata, int sec_no)
+static void cper_print_tstamp(const char *pfx,
+				   struct acpi_hest_generic_data_v300 *gdata)
+{
+	__u8 hour, min, sec, day, mon, year, century, *timestamp;
+
+	if (gdata->validation_bits & ACPI_HEST_GEN_VALID_TIMESTAMP) {
+		timestamp = (__u8 *)&(gdata->time_stamp);
+		sec       = bcd2bin(timestamp[0]);
+		min       = bcd2bin(timestamp[1]);
+		hour      = bcd2bin(timestamp[2]);
+		day       = bcd2bin(timestamp[4]);
+		mon       = bcd2bin(timestamp[5]);
+		year      = bcd2bin(timestamp[6]);
+		century   = bcd2bin(timestamp[7]);
+
+		printk("%s%ststamp: %02d%02d-%02d-%02d %02d:%02d:%02d\n", pfx,
+		       (timestamp[3] & 0x1 ? "precise " : "imprecise "),
+		       century, year, mon, day, hour, min, sec);
+	}
+}
+
+static void
+cper_estatus_print_section(const char *pfx, struct acpi_hest_generic_data *gdata,
+			   int sec_no)
 {
 	uuid_le *sec_type = (uuid_le *)gdata->section_type;
 	__u16 severity;
 	char newpfx[64];
 
+	if (acpi_hest_get_version(gdata) >= 3)
+		cper_print_tstamp(pfx, (struct acpi_hest_generic_data_v300 *)gdata);
+
 	severity = gdata->error_severity;
 	printk("%s""Error %d, type: %s\n", pfx, sec_no,
 	       cper_severity_str(severity));
@@ -403,14 +551,16 @@ static void cper_estatus_print_section(
 
 	snprintf(newpfx, sizeof(newpfx), "%s%s", pfx, INDENT_SP);
 	if (!uuid_le_cmp(*sec_type, CPER_SEC_PROC_GENERIC)) {
-		struct cper_sec_proc_generic *proc_err = (void *)(gdata + 1);
+		struct cper_sec_proc_generic *proc_err = acpi_hest_get_payload(gdata);
+
 		printk("%s""section_type: general processor error\n", newpfx);
 		if (gdata->error_data_length >= sizeof(*proc_err))
 			cper_print_proc_generic(newpfx, proc_err);
 		else
 			goto err_section_too_small;
 	} else if (!uuid_le_cmp(*sec_type, CPER_SEC_PLATFORM_MEM)) {
-		struct cper_sec_mem_err *mem_err = (void *)(gdata + 1);
+		struct cper_sec_mem_err *mem_err = acpi_hest_get_payload(gdata);
+
 		printk("%s""section_type: memory error\n", newpfx);
 		if (gdata->error_data_length >=
 		    sizeof(struct cper_sec_mem_err_old))
@@ -419,14 +569,32 @@ static void cper_estatus_print_section(
 		else
 			goto err_section_too_small;
 	} else if (!uuid_le_cmp(*sec_type, CPER_SEC_PCIE)) {
-		struct cper_sec_pcie *pcie = (void *)(gdata + 1);
+		struct cper_sec_pcie *pcie = acpi_hest_get_payload(gdata);
+
 		printk("%s""section_type: PCIe error\n", newpfx);
 		if (gdata->error_data_length >= sizeof(*pcie))
 			cper_print_pcie(newpfx, pcie, gdata);
 		else
 			goto err_section_too_small;
-	} else
-		printk("%s""section type: unknown, %pUl\n", newpfx, sec_type);
+#if defined(CONFIG_ARM64) || defined(CONFIG_ARM)
+	} else if (!uuid_le_cmp(*sec_type, CPER_SEC_PROC_ARM)) {
+		struct cper_sec_proc_arm *arm_err = acpi_hest_get_payload(gdata);
+
+		printk("%ssection_type: ARM processor error\n", newpfx);
+		if (gdata->error_data_length >= sizeof(*arm_err))
+			cper_print_proc_arm(newpfx, arm_err);
+		else
+			goto err_section_too_small;
+#endif
+	} else {
+		const void *err = acpi_hest_get_payload(gdata);
+
+		printk("%ssection type: unknown, %pUl\n", newpfx, sec_type);
+		printk("%ssection length: %#x\n", newpfx,
+		       gdata->error_data_length);
+		print_hex_dump(newpfx, "", DUMP_PREFIX_OFFSET, 16, 4, err,
+			       gdata->error_data_length, true);
+	}
 
 	return;
 
@@ -438,7 +606,7 @@ void cper_estatus_print(const char *pfx,
 			const struct acpi_hest_generic_status *estatus)
 {
 	struct acpi_hest_generic_data *gdata;
-	unsigned int data_len, gedata_len;
+	unsigned int data_len;
 	int sec_no = 0;
 	char newpfx[64];
 	__u16 severity;
@@ -452,11 +620,11 @@ void cper_estatus_print(const char *pfx,
 	data_len = estatus->data_length;
 	gdata = (struct acpi_hest_generic_data *)(estatus + 1);
 	snprintf(newpfx, sizeof(newpfx), "%s%s", pfx, INDENT_SP);
-	while (data_len >= sizeof(*gdata)) {
-		gedata_len = gdata->error_data_length;
+
+	while (data_len >= acpi_hest_get_size(gdata)) {
 		cper_estatus_print_section(newpfx, gdata, sec_no);
-		data_len -= gedata_len + sizeof(*gdata);
-		gdata = (void *)(gdata + 1) + gedata_len;
+		data_len -= acpi_hest_get_record_size(gdata);
+		gdata = acpi_hest_get_next(gdata);
 		sec_no++;
 	}
 }
@@ -486,12 +654,14 @@ int cper_estatus_check(const struct acpi_hest_generic_status *estatus)
 		return rc;
 	data_len = estatus->data_length;
 	gdata = (struct acpi_hest_generic_data *)(estatus + 1);
-	while (data_len >= sizeof(*gdata)) {
-		gedata_len = gdata->error_data_length;
-		if (gedata_len > data_len - sizeof(*gdata))
+
+	while (data_len >= acpi_hest_get_size(gdata)) {
+		gedata_len = acpi_hest_get_error_length(gdata);
+		if (gedata_len > data_len - acpi_hest_get_size(gdata))
 			return -EINVAL;
-		data_len -= gedata_len + sizeof(*gdata);
-		gdata = (void *)(gdata + 1) + gedata_len;
+
+		data_len -= acpi_hest_get_record_size(gdata);
+		gdata = acpi_hest_get_next(gdata);
 	}
 	if (data_len)
 		return -EINVAL;
diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig
index 676232a94f9f..f1fd5f44d1d4 100644
--- a/drivers/irqchip/Kconfig
+++ b/drivers/irqchip/Kconfig
@@ -39,7 +39,6 @@ config ARM_GIC_V3_ITS
 	bool
 	depends on PCI
 	depends on PCI_MSI
-	select ACPI_IORT if ACPI
 
 config ARM_NVIC
 	bool
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index aa587edaf9ea..e5197ffb7422 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -3,9 +3,10 @@
 #
 
 menu "Performance monitor support"
+	depends on PERF_EVENTS
 
 config ARM_PMU
-	depends on PERF_EVENTS && (ARM || ARM64)
+	depends on ARM || ARM64
 	bool "ARM PMU framework"
 	default y
 	help
@@ -18,7 +19,7 @@ config ARM_PMU_ACPI
 
 config QCOM_L2_PMU
 	bool "Qualcomm Technologies L2-cache PMU"
-	depends on ARCH_QCOM && ARM64 && PERF_EVENTS && ACPI
+	depends on ARCH_QCOM && ARM64 && ACPI
 	  help
 	  Provides support for the L2 cache performance monitor unit (PMU)
 	  in Qualcomm Technologies processors.
@@ -27,7 +28,7 @@ config QCOM_L2_PMU
 
 config QCOM_L3_PMU
 	bool "Qualcomm Technologies L3-cache PMU"
-	depends on ARCH_QCOM && ARM64 && PERF_EVENTS && ACPI
+	depends on ARCH_QCOM && ARM64 && ACPI
 	select QCOM_IRQ_COMBINER
 	help
 	   Provides support for the L3 cache performance monitor unit (PMU)
@@ -36,7 +37,7 @@ config QCOM_L3_PMU
 	   monitoring L3 cache events.
 
 config XGENE_PMU
-        depends on PERF_EVENTS && ARCH_XGENE
+        depends on ARCH_XGENE
         bool "APM X-Gene SoC PMU"
         default n
         help
diff --git a/drivers/perf/xgene_pmu.c b/drivers/perf/xgene_pmu.c
index 35b5289bc5da..e841282d690c 100644
--- a/drivers/perf/xgene_pmu.c
+++ b/drivers/perf/xgene_pmu.c
@@ -37,6 +37,8 @@
 
 #define CSW_CSWCR                       0x0000
 #define  CSW_CSWCR_DUALMCB_MASK         BIT(0)
+#define  CSW_CSWCR_MCB0_ROUTING(x)	(((x) & 0x0C) >> 2)
+#define  CSW_CSWCR_MCB1_ROUTING(x)	(((x) & 0x30) >> 4)
 #define MCBADDRMR                       0x0000
 #define  MCBADDRMR_DUALMCU_MODE_MASK    BIT(2)
 
@@ -50,8 +52,17 @@
 #define  PCPPMU_INT_L3C		BIT(2)
 #define  PCPPMU_INT_IOB		BIT(3)
 
+#define  PCPPMU_V3_INTMASK	0x00FF33FF
+#define  PCPPMU_V3_INTENMASK	0xFFFFFFFF
+#define  PCPPMU_V3_INTCLRMASK	0xFF00CC00
+#define  PCPPMU_V3_INT_MCU	0x000000FF
+#define  PCPPMU_V3_INT_MCB	0x00000300
+#define  PCPPMU_V3_INT_L3C	0x00FF0000
+#define  PCPPMU_V3_INT_IOB	0x00003000
+
 #define PMU_MAX_COUNTERS	4
-#define PMU_CNT_MAX_PERIOD	0x100000000ULL
+#define PMU_CNT_MAX_PERIOD	0xFFFFFFFFULL
+#define PMU_V3_CNT_MAX_PERIOD	0xFFFFFFFFFFFFFFFFULL
 #define PMU_OVERFLOW_MASK	0xF
 #define PMU_PMCR_E		BIT(0)
 #define PMU_PMCR_P		BIT(1)
@@ -73,6 +84,10 @@
 #define PMU_PMOVSR		0xC80
 #define PMU_PMCR		0xE04
 
+/* PMU registers for V3 */
+#define PMU_PMOVSCLR		0xC80
+#define PMU_PMOVSSET		0xCC0
+
 #define to_pmu_dev(p)     container_of(p, struct xgene_pmu_dev, pmu)
 #define GET_CNTR(ev)      (ev->hw.idx)
 #define GET_EVENTID(ev)   (ev->hw.config & 0xFFULL)
@@ -96,14 +111,33 @@ struct xgene_pmu_dev {
 	struct perf_event *pmu_counter_event[PMU_MAX_COUNTERS];
 };
 
+struct xgene_pmu_ops {
+	void (*mask_int)(struct xgene_pmu *pmu);
+	void (*unmask_int)(struct xgene_pmu *pmu);
+	u64 (*read_counter)(struct xgene_pmu_dev *pmu, int idx);
+	void (*write_counter)(struct xgene_pmu_dev *pmu, int idx, u64 val);
+	void (*write_evttype)(struct xgene_pmu_dev *pmu_dev, int idx, u32 val);
+	void (*write_agentmsk)(struct xgene_pmu_dev *pmu_dev, u32 val);
+	void (*write_agent1msk)(struct xgene_pmu_dev *pmu_dev, u32 val);
+	void (*enable_counter)(struct xgene_pmu_dev *pmu_dev, int idx);
+	void (*disable_counter)(struct xgene_pmu_dev *pmu_dev, int idx);
+	void (*enable_counter_int)(struct xgene_pmu_dev *pmu_dev, int idx);
+	void (*disable_counter_int)(struct xgene_pmu_dev *pmu_dev, int idx);
+	void (*reset_counters)(struct xgene_pmu_dev *pmu_dev);
+	void (*start_counters)(struct xgene_pmu_dev *pmu_dev);
+	void (*stop_counters)(struct xgene_pmu_dev *pmu_dev);
+};
+
 struct xgene_pmu {
 	struct device *dev;
 	int version;
 	void __iomem *pcppmu_csr;
 	u32 mcb_active_mask;
 	u32 mc_active_mask;
+	u32 l3c_active_mask;
 	cpumask_t cpu;
 	raw_spinlock_t lock;
+	const struct xgene_pmu_ops *ops;
 	struct list_head l3cpmus;
 	struct list_head iobpmus;
 	struct list_head mcbpmus;
@@ -125,11 +159,13 @@ struct xgene_pmu_data {
 enum xgene_pmu_version {
 	PCP_PMU_V1 = 1,
 	PCP_PMU_V2,
+	PCP_PMU_V3,
 };
 
 enum xgene_pmu_dev_type {
 	PMU_TYPE_L3C = 0,
 	PMU_TYPE_IOB,
+	PMU_TYPE_IOB_SLOW,
 	PMU_TYPE_MCB,
 	PMU_TYPE_MC,
 };
@@ -195,6 +231,56 @@ static const struct attribute_group mc_pmu_format_attr_group = {
 	.attrs = mc_pmu_format_attrs,
 };
 
+static struct attribute *l3c_pmu_v3_format_attrs[] = {
+	XGENE_PMU_FORMAT_ATTR(l3c_eventid, "config:0-39"),
+	NULL,
+};
+
+static struct attribute *iob_pmu_v3_format_attrs[] = {
+	XGENE_PMU_FORMAT_ATTR(iob_eventid, "config:0-47"),
+	NULL,
+};
+
+static struct attribute *iob_slow_pmu_v3_format_attrs[] = {
+	XGENE_PMU_FORMAT_ATTR(iob_slow_eventid, "config:0-16"),
+	NULL,
+};
+
+static struct attribute *mcb_pmu_v3_format_attrs[] = {
+	XGENE_PMU_FORMAT_ATTR(mcb_eventid, "config:0-35"),
+	NULL,
+};
+
+static struct attribute *mc_pmu_v3_format_attrs[] = {
+	XGENE_PMU_FORMAT_ATTR(mc_eventid, "config:0-44"),
+	NULL,
+};
+
+static const struct attribute_group l3c_pmu_v3_format_attr_group = {
+	.name = "format",
+	.attrs = l3c_pmu_v3_format_attrs,
+};
+
+static const struct attribute_group iob_pmu_v3_format_attr_group = {
+	.name = "format",
+	.attrs = iob_pmu_v3_format_attrs,
+};
+
+static const struct attribute_group iob_slow_pmu_v3_format_attr_group = {
+	.name = "format",
+	.attrs = iob_slow_pmu_v3_format_attrs,
+};
+
+static const struct attribute_group mcb_pmu_v3_format_attr_group = {
+	.name = "format",
+	.attrs = mcb_pmu_v3_format_attrs,
+};
+
+static const struct attribute_group mc_pmu_v3_format_attr_group = {
+	.name = "format",
+	.attrs = mc_pmu_v3_format_attrs,
+};
+
 /*
  * sysfs event attributes
  */
@@ -311,6 +397,219 @@ static const struct attribute_group mc_pmu_events_attr_group = {
 	.attrs = mc_pmu_events_attrs,
 };
 
+static struct attribute *l3c_pmu_v3_events_attrs[] = {
+	XGENE_PMU_EVENT_ATTR(cycle-count,			0x00),
+	XGENE_PMU_EVENT_ATTR(read-hit,				0x01),
+	XGENE_PMU_EVENT_ATTR(read-miss,				0x02),
+	XGENE_PMU_EVENT_ATTR(index-flush-eviction,		0x03),
+	XGENE_PMU_EVENT_ATTR(write-caused-replacement,		0x04),
+	XGENE_PMU_EVENT_ATTR(write-not-caused-replacement,	0x05),
+	XGENE_PMU_EVENT_ATTR(clean-eviction,			0x06),
+	XGENE_PMU_EVENT_ATTR(dirty-eviction,			0x07),
+	XGENE_PMU_EVENT_ATTR(read,				0x08),
+	XGENE_PMU_EVENT_ATTR(write,				0x09),
+	XGENE_PMU_EVENT_ATTR(request,				0x0a),
+	XGENE_PMU_EVENT_ATTR(tq-bank-conflict-issue-stall,	0x0b),
+	XGENE_PMU_EVENT_ATTR(tq-full,				0x0c),
+	XGENE_PMU_EVENT_ATTR(ackq-full,				0x0d),
+	XGENE_PMU_EVENT_ATTR(wdb-full,				0x0e),
+	XGENE_PMU_EVENT_ATTR(odb-full,				0x10),
+	XGENE_PMU_EVENT_ATTR(wbq-full,				0x11),
+	XGENE_PMU_EVENT_ATTR(input-req-async-fifo-stall,	0x12),
+	XGENE_PMU_EVENT_ATTR(output-req-async-fifo-stall,	0x13),
+	XGENE_PMU_EVENT_ATTR(output-data-async-fifo-stall,	0x14),
+	XGENE_PMU_EVENT_ATTR(total-insertion,			0x15),
+	XGENE_PMU_EVENT_ATTR(sip-insertions-r-set,		0x16),
+	XGENE_PMU_EVENT_ATTR(sip-insertions-r-clear,		0x17),
+	XGENE_PMU_EVENT_ATTR(dip-insertions-r-set,		0x18),
+	XGENE_PMU_EVENT_ATTR(dip-insertions-r-clear,		0x19),
+	XGENE_PMU_EVENT_ATTR(dip-insertions-force-r-set,	0x1a),
+	XGENE_PMU_EVENT_ATTR(egression,				0x1b),
+	XGENE_PMU_EVENT_ATTR(replacement,			0x1c),
+	XGENE_PMU_EVENT_ATTR(old-replacement,			0x1d),
+	XGENE_PMU_EVENT_ATTR(young-replacement,			0x1e),
+	XGENE_PMU_EVENT_ATTR(r-set-replacement,			0x1f),
+	XGENE_PMU_EVENT_ATTR(r-clear-replacement,		0x20),
+	XGENE_PMU_EVENT_ATTR(old-r-replacement,			0x21),
+	XGENE_PMU_EVENT_ATTR(old-nr-replacement,		0x22),
+	XGENE_PMU_EVENT_ATTR(young-r-replacement,		0x23),
+	XGENE_PMU_EVENT_ATTR(young-nr-replacement,		0x24),
+	XGENE_PMU_EVENT_ATTR(bloomfilter-clearing,		0x25),
+	XGENE_PMU_EVENT_ATTR(generation-flip,			0x26),
+	XGENE_PMU_EVENT_ATTR(vcc-droop-detected,		0x27),
+	NULL,
+};
+
+static struct attribute *iob_fast_pmu_v3_events_attrs[] = {
+	XGENE_PMU_EVENT_ATTR(cycle-count,			0x00),
+	XGENE_PMU_EVENT_ATTR(pa-req-buf-alloc-all,		0x01),
+	XGENE_PMU_EVENT_ATTR(pa-req-buf-alloc-rd,		0x02),
+	XGENE_PMU_EVENT_ATTR(pa-req-buf-alloc-wr,		0x03),
+	XGENE_PMU_EVENT_ATTR(pa-all-cp-req,			0x04),
+	XGENE_PMU_EVENT_ATTR(pa-cp-blk-req,			0x05),
+	XGENE_PMU_EVENT_ATTR(pa-cp-ptl-req,			0x06),
+	XGENE_PMU_EVENT_ATTR(pa-cp-rd-req,			0x07),
+	XGENE_PMU_EVENT_ATTR(pa-cp-wr-req,			0x08),
+	XGENE_PMU_EVENT_ATTR(ba-all-req,			0x09),
+	XGENE_PMU_EVENT_ATTR(ba-rd-req,				0x0a),
+	XGENE_PMU_EVENT_ATTR(ba-wr-req,				0x0b),
+	XGENE_PMU_EVENT_ATTR(pa-rd-shared-req-issued,		0x10),
+	XGENE_PMU_EVENT_ATTR(pa-rd-exclusive-req-issued,	0x11),
+	XGENE_PMU_EVENT_ATTR(pa-wr-invalidate-req-issued-stashable, 0x12),
+	XGENE_PMU_EVENT_ATTR(pa-wr-invalidate-req-issued-nonstashable, 0x13),
+	XGENE_PMU_EVENT_ATTR(pa-wr-back-req-issued-stashable,	0x14),
+	XGENE_PMU_EVENT_ATTR(pa-wr-back-req-issued-nonstashable, 0x15),
+	XGENE_PMU_EVENT_ATTR(pa-ptl-wr-req,			0x16),
+	XGENE_PMU_EVENT_ATTR(pa-ptl-rd-req,			0x17),
+	XGENE_PMU_EVENT_ATTR(pa-wr-back-clean-data,		0x18),
+	XGENE_PMU_EVENT_ATTR(pa-wr-back-cancelled-on-SS,	0x1b),
+	XGENE_PMU_EVENT_ATTR(pa-barrier-occurrence,		0x1c),
+	XGENE_PMU_EVENT_ATTR(pa-barrier-cycles,			0x1d),
+	XGENE_PMU_EVENT_ATTR(pa-total-cp-snoops,		0x20),
+	XGENE_PMU_EVENT_ATTR(pa-rd-shared-snoop,		0x21),
+	XGENE_PMU_EVENT_ATTR(pa-rd-shared-snoop-hit,		0x22),
+	XGENE_PMU_EVENT_ATTR(pa-rd-exclusive-snoop,		0x23),
+	XGENE_PMU_EVENT_ATTR(pa-rd-exclusive-snoop-hit,		0x24),
+	XGENE_PMU_EVENT_ATTR(pa-rd-wr-invalid-snoop,		0x25),
+	XGENE_PMU_EVENT_ATTR(pa-rd-wr-invalid-snoop-hit,	0x26),
+	XGENE_PMU_EVENT_ATTR(pa-req-buffer-full,		0x28),
+	XGENE_PMU_EVENT_ATTR(cswlf-outbound-req-fifo-full,	0x29),
+	XGENE_PMU_EVENT_ATTR(cswlf-inbound-snoop-fifo-backpressure, 0x2a),
+	XGENE_PMU_EVENT_ATTR(cswlf-outbound-lack-fifo-full,	0x2b),
+	XGENE_PMU_EVENT_ATTR(cswlf-inbound-gack-fifo-backpressure, 0x2c),
+	XGENE_PMU_EVENT_ATTR(cswlf-outbound-data-fifo-full,	0x2d),
+	XGENE_PMU_EVENT_ATTR(cswlf-inbound-data-fifo-backpressure, 0x2e),
+	XGENE_PMU_EVENT_ATTR(cswlf-inbound-req-backpressure,	0x2f),
+	NULL,
+};
+
+static struct attribute *iob_slow_pmu_v3_events_attrs[] = {
+	XGENE_PMU_EVENT_ATTR(cycle-count,			0x00),
+	XGENE_PMU_EVENT_ATTR(pa-axi0-rd-req,			0x01),
+	XGENE_PMU_EVENT_ATTR(pa-axi0-wr-req,			0x02),
+	XGENE_PMU_EVENT_ATTR(pa-axi1-rd-req,			0x03),
+	XGENE_PMU_EVENT_ATTR(pa-axi1-wr-req,			0x04),
+	XGENE_PMU_EVENT_ATTR(ba-all-axi-req,			0x07),
+	XGENE_PMU_EVENT_ATTR(ba-axi-rd-req,			0x08),
+	XGENE_PMU_EVENT_ATTR(ba-axi-wr-req,			0x09),
+	XGENE_PMU_EVENT_ATTR(ba-free-list-empty,		0x10),
+	NULL,
+};
+
+static struct attribute *mcb_pmu_v3_events_attrs[] = {
+	XGENE_PMU_EVENT_ATTR(cycle-count,			0x00),
+	XGENE_PMU_EVENT_ATTR(req-receive,			0x01),
+	XGENE_PMU_EVENT_ATTR(rd-req-recv,			0x02),
+	XGENE_PMU_EVENT_ATTR(rd-req-recv-2,			0x03),
+	XGENE_PMU_EVENT_ATTR(wr-req-recv,			0x04),
+	XGENE_PMU_EVENT_ATTR(wr-req-recv-2,			0x05),
+	XGENE_PMU_EVENT_ATTR(rd-req-sent-to-mcu,		0x06),
+	XGENE_PMU_EVENT_ATTR(rd-req-sent-to-mcu-2,		0x07),
+	XGENE_PMU_EVENT_ATTR(rd-req-sent-to-spec-mcu,		0x08),
+	XGENE_PMU_EVENT_ATTR(rd-req-sent-to-spec-mcu-2,		0x09),
+	XGENE_PMU_EVENT_ATTR(glbl-ack-recv-for-rd-sent-to-spec-mcu, 0x0a),
+	XGENE_PMU_EVENT_ATTR(glbl-ack-go-recv-for-rd-sent-to-spec-mcu, 0x0b),
+	XGENE_PMU_EVENT_ATTR(glbl-ack-nogo-recv-for-rd-sent-to-spec-mcu, 0x0c),
+	XGENE_PMU_EVENT_ATTR(glbl-ack-go-recv-any-rd-req,	0x0d),
+	XGENE_PMU_EVENT_ATTR(glbl-ack-go-recv-any-rd-req-2,	0x0e),
+	XGENE_PMU_EVENT_ATTR(wr-req-sent-to-mcu,		0x0f),
+	XGENE_PMU_EVENT_ATTR(gack-recv,				0x10),
+	XGENE_PMU_EVENT_ATTR(rd-gack-recv,			0x11),
+	XGENE_PMU_EVENT_ATTR(wr-gack-recv,			0x12),
+	XGENE_PMU_EVENT_ATTR(cancel-rd-gack,			0x13),
+	XGENE_PMU_EVENT_ATTR(cancel-wr-gack,			0x14),
+	XGENE_PMU_EVENT_ATTR(mcb-csw-req-stall,			0x15),
+	XGENE_PMU_EVENT_ATTR(mcu-req-intf-blocked,		0x16),
+	XGENE_PMU_EVENT_ATTR(mcb-mcu-rd-intf-stall,		0x17),
+	XGENE_PMU_EVENT_ATTR(csw-rd-intf-blocked,		0x18),
+	XGENE_PMU_EVENT_ATTR(csw-local-ack-intf-blocked,	0x19),
+	XGENE_PMU_EVENT_ATTR(mcu-req-table-full,		0x1a),
+	XGENE_PMU_EVENT_ATTR(mcu-stat-table-full,		0x1b),
+	XGENE_PMU_EVENT_ATTR(mcu-wr-table-full,			0x1c),
+	XGENE_PMU_EVENT_ATTR(mcu-rdreceipt-resp,		0x1d),
+	XGENE_PMU_EVENT_ATTR(mcu-wrcomplete-resp,		0x1e),
+	XGENE_PMU_EVENT_ATTR(mcu-retryack-resp,			0x1f),
+	XGENE_PMU_EVENT_ATTR(mcu-pcrdgrant-resp,		0x20),
+	XGENE_PMU_EVENT_ATTR(mcu-req-from-lastload,		0x21),
+	XGENE_PMU_EVENT_ATTR(mcu-req-from-bypass,		0x22),
+	XGENE_PMU_EVENT_ATTR(volt-droop-detect,			0x23),
+	NULL,
+};
+
+static struct attribute *mc_pmu_v3_events_attrs[] = {
+	XGENE_PMU_EVENT_ATTR(cycle-count,			0x00),
+	XGENE_PMU_EVENT_ATTR(act-sent,				0x01),
+	XGENE_PMU_EVENT_ATTR(pre-sent,				0x02),
+	XGENE_PMU_EVENT_ATTR(rd-sent,				0x03),
+	XGENE_PMU_EVENT_ATTR(rda-sent,				0x04),
+	XGENE_PMU_EVENT_ATTR(wr-sent,				0x05),
+	XGENE_PMU_EVENT_ATTR(wra-sent,				0x06),
+	XGENE_PMU_EVENT_ATTR(pd-entry-vld,			0x07),
+	XGENE_PMU_EVENT_ATTR(sref-entry-vld,			0x08),
+	XGENE_PMU_EVENT_ATTR(prea-sent,				0x09),
+	XGENE_PMU_EVENT_ATTR(ref-sent,				0x0a),
+	XGENE_PMU_EVENT_ATTR(rd-rda-sent,			0x0b),
+	XGENE_PMU_EVENT_ATTR(wr-wra-sent,			0x0c),
+	XGENE_PMU_EVENT_ATTR(raw-hazard,			0x0d),
+	XGENE_PMU_EVENT_ATTR(war-hazard,			0x0e),
+	XGENE_PMU_EVENT_ATTR(waw-hazard,			0x0f),
+	XGENE_PMU_EVENT_ATTR(rar-hazard,			0x10),
+	XGENE_PMU_EVENT_ATTR(raw-war-waw-hazard,		0x11),
+	XGENE_PMU_EVENT_ATTR(hprd-lprd-wr-req-vld,		0x12),
+	XGENE_PMU_EVENT_ATTR(lprd-req-vld,			0x13),
+	XGENE_PMU_EVENT_ATTR(hprd-req-vld,			0x14),
+	XGENE_PMU_EVENT_ATTR(hprd-lprd-req-vld,			0x15),
+	XGENE_PMU_EVENT_ATTR(wr-req-vld,			0x16),
+	XGENE_PMU_EVENT_ATTR(partial-wr-req-vld,		0x17),
+	XGENE_PMU_EVENT_ATTR(rd-retry,				0x18),
+	XGENE_PMU_EVENT_ATTR(wr-retry,				0x19),
+	XGENE_PMU_EVENT_ATTR(retry-gnt,				0x1a),
+	XGENE_PMU_EVENT_ATTR(rank-change,			0x1b),
+	XGENE_PMU_EVENT_ATTR(dir-change,			0x1c),
+	XGENE_PMU_EVENT_ATTR(rank-dir-change,			0x1d),
+	XGENE_PMU_EVENT_ATTR(rank-active,			0x1e),
+	XGENE_PMU_EVENT_ATTR(rank-idle,				0x1f),
+	XGENE_PMU_EVENT_ATTR(rank-pd,				0x20),
+	XGENE_PMU_EVENT_ATTR(rank-sref,				0x21),
+	XGENE_PMU_EVENT_ATTR(queue-fill-gt-thresh,		0x22),
+	XGENE_PMU_EVENT_ATTR(queue-rds-gt-thresh,		0x23),
+	XGENE_PMU_EVENT_ATTR(queue-wrs-gt-thresh,		0x24),
+	XGENE_PMU_EVENT_ATTR(phy-updt-complt,			0x25),
+	XGENE_PMU_EVENT_ATTR(tz-fail,				0x26),
+	XGENE_PMU_EVENT_ATTR(dram-errc,				0x27),
+	XGENE_PMU_EVENT_ATTR(dram-errd,				0x28),
+	XGENE_PMU_EVENT_ATTR(rd-enq,				0x29),
+	XGENE_PMU_EVENT_ATTR(wr-enq,				0x2a),
+	XGENE_PMU_EVENT_ATTR(tmac-limit-reached,		0x2b),
+	XGENE_PMU_EVENT_ATTR(tmaw-tracker-full,			0x2c),
+	NULL,
+};
+
+static const struct attribute_group l3c_pmu_v3_events_attr_group = {
+	.name = "events",
+	.attrs = l3c_pmu_v3_events_attrs,
+};
+
+static const struct attribute_group iob_fast_pmu_v3_events_attr_group = {
+	.name = "events",
+	.attrs = iob_fast_pmu_v3_events_attrs,
+};
+
+static const struct attribute_group iob_slow_pmu_v3_events_attr_group = {
+	.name = "events",
+	.attrs = iob_slow_pmu_v3_events_attrs,
+};
+
+static const struct attribute_group mcb_pmu_v3_events_attr_group = {
+	.name = "events",
+	.attrs = mcb_pmu_v3_events_attrs,
+};
+
+static const struct attribute_group mc_pmu_v3_events_attr_group = {
+	.name = "events",
+	.attrs = mc_pmu_v3_events_attrs,
+};
+
 /*
  * sysfs cpumask attributes
  */
@@ -334,7 +633,7 @@ static const struct attribute_group pmu_cpumask_attr_group = {
 };
 
 /*
- * Per PMU device attribute groups
+ * Per PMU device attribute groups of PMU v1 and v2
  */
 static const struct attribute_group *l3c_pmu_attr_groups[] = {
 	&l3c_pmu_format_attr_group,
@@ -364,6 +663,44 @@ static const struct attribute_group *mc_pmu_attr_groups[] = {
 	NULL
 };
 
+/*
+ * Per PMU device attribute groups of PMU v3
+ */
+static const struct attribute_group *l3c_pmu_v3_attr_groups[] = {
+	&l3c_pmu_v3_format_attr_group,
+	&pmu_cpumask_attr_group,
+	&l3c_pmu_v3_events_attr_group,
+	NULL
+};
+
+static const struct attribute_group *iob_fast_pmu_v3_attr_groups[] = {
+	&iob_pmu_v3_format_attr_group,
+	&pmu_cpumask_attr_group,
+	&iob_fast_pmu_v3_events_attr_group,
+	NULL
+};
+
+static const struct attribute_group *iob_slow_pmu_v3_attr_groups[] = {
+	&iob_slow_pmu_v3_format_attr_group,
+	&pmu_cpumask_attr_group,
+	&iob_slow_pmu_v3_events_attr_group,
+	NULL
+};
+
+static const struct attribute_group *mcb_pmu_v3_attr_groups[] = {
+	&mcb_pmu_v3_format_attr_group,
+	&pmu_cpumask_attr_group,
+	&mcb_pmu_v3_events_attr_group,
+	NULL
+};
+
+static const struct attribute_group *mc_pmu_v3_attr_groups[] = {
+	&mc_pmu_v3_format_attr_group,
+	&pmu_cpumask_attr_group,
+	&mc_pmu_v3_events_attr_group,
+	NULL
+};
+
 static int get_next_avail_cntr(struct xgene_pmu_dev *pmu_dev)
 {
 	int cntr;
@@ -387,23 +724,67 @@ static inline void xgene_pmu_mask_int(struct xgene_pmu *xgene_pmu)
 	writel(PCPPMU_INTENMASK, xgene_pmu->pcppmu_csr + PCPPMU_INTMASK_REG);
 }
 
+static inline void xgene_pmu_v3_mask_int(struct xgene_pmu *xgene_pmu)
+{
+	writel(PCPPMU_V3_INTENMASK, xgene_pmu->pcppmu_csr + PCPPMU_INTMASK_REG);
+}
+
 static inline void xgene_pmu_unmask_int(struct xgene_pmu *xgene_pmu)
 {
 	writel(PCPPMU_INTCLRMASK, xgene_pmu->pcppmu_csr + PCPPMU_INTMASK_REG);
 }
 
-static inline u32 xgene_pmu_read_counter(struct xgene_pmu_dev *pmu_dev, int idx)
+static inline void xgene_pmu_v3_unmask_int(struct xgene_pmu *xgene_pmu)
+{
+	writel(PCPPMU_V3_INTCLRMASK,
+	       xgene_pmu->pcppmu_csr + PCPPMU_INTMASK_REG);
+}
+
+static inline u64 xgene_pmu_read_counter32(struct xgene_pmu_dev *pmu_dev,
+					   int idx)
 {
 	return readl(pmu_dev->inf->csr + PMU_PMEVCNTR0 + (4 * idx));
 }
 
+static inline u64 xgene_pmu_read_counter64(struct xgene_pmu_dev *pmu_dev,
+					   int idx)
+{
+	u32 lo, hi;
+
+	/*
+	 * v3 has 64-bit counter registers composed by 2 32-bit registers
+	 * This can be a problem if the counter increases and carries
+	 * out of bit [31] between 2 reads. The extra reads would help
+	 * to prevent this issue.
+	 */
+	do {
+		hi = xgene_pmu_read_counter32(pmu_dev, 2 * idx + 1);
+		lo = xgene_pmu_read_counter32(pmu_dev, 2 * idx);
+	} while (hi != xgene_pmu_read_counter32(pmu_dev, 2 * idx + 1));
+
+	return (((u64)hi << 32) | lo);
+}
+
 static inline void
-xgene_pmu_write_counter(struct xgene_pmu_dev *pmu_dev, int idx, u32 val)
+xgene_pmu_write_counter32(struct xgene_pmu_dev *pmu_dev, int idx, u64 val)
 {
 	writel(val, pmu_dev->inf->csr + PMU_PMEVCNTR0 + (4 * idx));
 }
 
 static inline void
+xgene_pmu_write_counter64(struct xgene_pmu_dev *pmu_dev, int idx, u64 val)
+{
+	u32 cnt_lo, cnt_hi;
+
+	cnt_hi = upper_32_bits(val);
+	cnt_lo = lower_32_bits(val);
+
+	/* v3 has 64-bit counter registers composed by 2 32-bit registers */
+	xgene_pmu_write_counter32(pmu_dev, 2 * idx, cnt_lo);
+	xgene_pmu_write_counter32(pmu_dev, 2 * idx + 1, cnt_hi);
+}
+
+static inline void
 xgene_pmu_write_evttype(struct xgene_pmu_dev *pmu_dev, int idx, u32 val)
 {
 	writel(val, pmu_dev->inf->csr + PMU_PMEVTYPER0 + (4 * idx));
@@ -416,12 +797,18 @@ xgene_pmu_write_agentmsk(struct xgene_pmu_dev *pmu_dev, u32 val)
 }
 
 static inline void
+xgene_pmu_v3_write_agentmsk(struct xgene_pmu_dev *pmu_dev, u32 val) { }
+
+static inline void
 xgene_pmu_write_agent1msk(struct xgene_pmu_dev *pmu_dev, u32 val)
 {
 	writel(val, pmu_dev->inf->csr + PMU_PMAMR1);
 }
 
 static inline void
+xgene_pmu_v3_write_agent1msk(struct xgene_pmu_dev *pmu_dev, u32 val) { }
+
+static inline void
 xgene_pmu_enable_counter(struct xgene_pmu_dev *pmu_dev, int idx)
 {
 	u32 val;
@@ -491,20 +878,22 @@ static inline void xgene_pmu_stop_counters(struct xgene_pmu_dev *pmu_dev)
 static void xgene_perf_pmu_enable(struct pmu *pmu)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 	int enabled = bitmap_weight(pmu_dev->cntr_assign_mask,
 			pmu_dev->max_counters);
 
 	if (!enabled)
 		return;
 
-	xgene_pmu_start_counters(pmu_dev);
+	xgene_pmu->ops->start_counters(pmu_dev);
 }
 
 static void xgene_perf_pmu_disable(struct pmu *pmu)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 
-	xgene_pmu_stop_counters(pmu_dev);
+	xgene_pmu->ops->stop_counters(pmu_dev);
 }
 
 static int xgene_perf_event_init(struct perf_event *event)
@@ -572,49 +961,56 @@ static int xgene_perf_event_init(struct perf_event *event)
 static void xgene_perf_enable_event(struct perf_event *event)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(event->pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 
-	xgene_pmu_write_evttype(pmu_dev, GET_CNTR(event), GET_EVENTID(event));
-	xgene_pmu_write_agentmsk(pmu_dev, ~((u32)GET_AGENTID(event)));
+	xgene_pmu->ops->write_evttype(pmu_dev, GET_CNTR(event),
+				      GET_EVENTID(event));
+	xgene_pmu->ops->write_agentmsk(pmu_dev, ~((u32)GET_AGENTID(event)));
 	if (pmu_dev->inf->type == PMU_TYPE_IOB)
-		xgene_pmu_write_agent1msk(pmu_dev, ~((u32)GET_AGENT1ID(event)));
+		xgene_pmu->ops->write_agent1msk(pmu_dev,
+						~((u32)GET_AGENT1ID(event)));
 
-	xgene_pmu_enable_counter(pmu_dev, GET_CNTR(event));
-	xgene_pmu_enable_counter_int(pmu_dev, GET_CNTR(event));
+	xgene_pmu->ops->enable_counter(pmu_dev, GET_CNTR(event));
+	xgene_pmu->ops->enable_counter_int(pmu_dev, GET_CNTR(event));
 }
 
 static void xgene_perf_disable_event(struct perf_event *event)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(event->pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 
-	xgene_pmu_disable_counter(pmu_dev, GET_CNTR(event));
-	xgene_pmu_disable_counter_int(pmu_dev, GET_CNTR(event));
+	xgene_pmu->ops->disable_counter(pmu_dev, GET_CNTR(event));
+	xgene_pmu->ops->disable_counter_int(pmu_dev, GET_CNTR(event));
 }
 
 static void xgene_perf_event_set_period(struct perf_event *event)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(event->pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 	struct hw_perf_event *hw = &event->hw;
 	/*
-	 * The X-Gene PMU counters have a period of 2^32. To account for the
-	 * possiblity of extreme interrupt latency we program for a period of
-	 * half that. Hopefully we can handle the interrupt before another 2^31
+	 * For 32 bit counter, it has a period of 2^32. To account for the
+	 * possibility of extreme interrupt latency we program for a period of
+	 * half that. Hopefully, we can handle the interrupt before another 2^31
 	 * events occur and the counter overtakes its previous value.
+	 * For 64 bit counter, we don't expect it overflow.
 	 */
 	u64 val = 1ULL << 31;
 
 	local64_set(&hw->prev_count, val);
-	xgene_pmu_write_counter(pmu_dev, hw->idx, (u32) val);
+	xgene_pmu->ops->write_counter(pmu_dev, hw->idx, val);
 }
 
 static void xgene_perf_event_update(struct perf_event *event)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(event->pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 	struct hw_perf_event *hw = &event->hw;
 	u64 delta, prev_raw_count, new_raw_count;
 
 again:
 	prev_raw_count = local64_read(&hw->prev_count);
-	new_raw_count = xgene_pmu_read_counter(pmu_dev, GET_CNTR(event));
+	new_raw_count = xgene_pmu->ops->read_counter(pmu_dev, GET_CNTR(event));
 
 	if (local64_cmpxchg(&hw->prev_count, prev_raw_count,
 			    new_raw_count) != prev_raw_count)
@@ -633,6 +1029,7 @@ static void xgene_perf_read(struct perf_event *event)
 static void xgene_perf_start(struct perf_event *event, int flags)
 {
 	struct xgene_pmu_dev *pmu_dev = to_pmu_dev(event->pmu);
+	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
 	struct hw_perf_event *hw = &event->hw;
 
 	if (WARN_ON_ONCE(!(hw->state & PERF_HES_STOPPED)))
@@ -646,8 +1043,8 @@ static void xgene_perf_start(struct perf_event *event, int flags)
 	if (flags & PERF_EF_RELOAD) {
 		u64 prev_raw_count =  local64_read(&hw->prev_count);
 
-		xgene_pmu_write_counter(pmu_dev, GET_CNTR(event),
-					(u32) prev_raw_count);
+		xgene_pmu->ops->write_counter(pmu_dev, GET_CNTR(event),
+					      prev_raw_count);
 	}
 
 	xgene_perf_enable_event(event);
@@ -713,7 +1110,10 @@ static int xgene_init_perf(struct xgene_pmu_dev *pmu_dev, char *name)
 {
 	struct xgene_pmu *xgene_pmu;
 
-	pmu_dev->max_period = PMU_CNT_MAX_PERIOD - 1;
+	if (pmu_dev->parent->version == PCP_PMU_V3)
+		pmu_dev->max_period = PMU_V3_CNT_MAX_PERIOD;
+	else
+		pmu_dev->max_period = PMU_CNT_MAX_PERIOD;
 	/* First version PMU supports only single event counter */
 	xgene_pmu = pmu_dev->parent;
 	if (xgene_pmu->version == PCP_PMU_V1)
@@ -736,8 +1136,8 @@ static int xgene_init_perf(struct xgene_pmu_dev *pmu_dev, char *name)
 	};
 
 	/* Hardware counter init */
-	xgene_pmu_stop_counters(pmu_dev);
-	xgene_pmu_reset_counters(pmu_dev);
+	xgene_pmu->ops->stop_counters(pmu_dev);
+	xgene_pmu->ops->reset_counters(pmu_dev);
 
 	return perf_pmu_register(&pmu_dev->pmu, name, -1);
 }
@@ -758,20 +1158,38 @@ xgene_pmu_dev_add(struct xgene_pmu *xgene_pmu, struct xgene_pmu_dev_ctx *ctx)
 
 	switch (pmu->inf->type) {
 	case PMU_TYPE_L3C:
-		pmu->attr_groups = l3c_pmu_attr_groups;
+		if (!(xgene_pmu->l3c_active_mask & pmu->inf->enable_mask))
+			goto dev_err;
+		if (xgene_pmu->version == PCP_PMU_V3)
+			pmu->attr_groups = l3c_pmu_v3_attr_groups;
+		else
+			pmu->attr_groups = l3c_pmu_attr_groups;
 		break;
 	case PMU_TYPE_IOB:
-		pmu->attr_groups = iob_pmu_attr_groups;
+		if (xgene_pmu->version == PCP_PMU_V3)
+			pmu->attr_groups = iob_fast_pmu_v3_attr_groups;
+		else
+			pmu->attr_groups = iob_pmu_attr_groups;
+		break;
+	case PMU_TYPE_IOB_SLOW:
+		if (xgene_pmu->version == PCP_PMU_V3)
+			pmu->attr_groups = iob_slow_pmu_v3_attr_groups;
 		break;
 	case PMU_TYPE_MCB:
 		if (!(xgene_pmu->mcb_active_mask & pmu->inf->enable_mask))
 			goto dev_err;
-		pmu->attr_groups = mcb_pmu_attr_groups;
+		if (xgene_pmu->version == PCP_PMU_V3)
+			pmu->attr_groups = mcb_pmu_v3_attr_groups;
+		else
+			pmu->attr_groups = mcb_pmu_attr_groups;
 		break;
 	case PMU_TYPE_MC:
 		if (!(xgene_pmu->mc_active_mask & pmu->inf->enable_mask))
 			goto dev_err;
-		pmu->attr_groups = mc_pmu_attr_groups;
+		if (xgene_pmu->version == PCP_PMU_V3)
+			pmu->attr_groups = mc_pmu_v3_attr_groups;
+		else
+			pmu->attr_groups = mc_pmu_attr_groups;
 		break;
 	default:
 		return -EINVAL;
@@ -795,18 +1213,27 @@ dev_err:
 static void _xgene_pmu_isr(int irq, struct xgene_pmu_dev *pmu_dev)
 {
 	struct xgene_pmu *xgene_pmu = pmu_dev->parent;
+	void __iomem *csr = pmu_dev->inf->csr;
 	u32 pmovsr;
 	int idx;
 
-	pmovsr = readl(pmu_dev->inf->csr + PMU_PMOVSR) & PMU_OVERFLOW_MASK;
+	xgene_pmu->ops->stop_counters(pmu_dev);
+
+	if (xgene_pmu->version == PCP_PMU_V3)
+		pmovsr = readl(csr + PMU_PMOVSSET) & PMU_OVERFLOW_MASK;
+	else
+		pmovsr = readl(csr + PMU_PMOVSR) & PMU_OVERFLOW_MASK;
+
 	if (!pmovsr)
-		return;
+		goto out;
 
 	/* Clear interrupt flag */
 	if (xgene_pmu->version == PCP_PMU_V1)
-		writel(0x0, pmu_dev->inf->csr + PMU_PMOVSR);
+		writel(0x0, csr + PMU_PMOVSR);
+	else if (xgene_pmu->version == PCP_PMU_V2)
+		writel(pmovsr, csr + PMU_PMOVSR);
 	else
-		writel(pmovsr, pmu_dev->inf->csr + PMU_PMOVSR);
+		writel(pmovsr, csr + PMU_PMOVSCLR);
 
 	for (idx = 0; idx < PMU_MAX_COUNTERS; idx++) {
 		struct perf_event *event = pmu_dev->pmu_counter_event[idx];
@@ -818,10 +1245,14 @@ static void _xgene_pmu_isr(int irq, struct xgene_pmu_dev *pmu_dev)
 		xgene_perf_event_update(event);
 		xgene_perf_event_set_period(event);
 	}
+
+out:
+	xgene_pmu->ops->start_counters(pmu_dev);
 }
 
 static irqreturn_t xgene_pmu_isr(int irq, void *dev_id)
 {
+	u32 intr_mcu, intr_mcb, intr_l3c, intr_iob;
 	struct xgene_pmu_dev_ctx *ctx;
 	struct xgene_pmu *xgene_pmu = dev_id;
 	unsigned long flags;
@@ -831,22 +1262,33 @@ static irqreturn_t xgene_pmu_isr(int irq, void *dev_id)
 
 	/* Get Interrupt PMU source */
 	val = readl(xgene_pmu->pcppmu_csr + PCPPMU_INTSTATUS_REG);
-	if (val & PCPPMU_INT_MCU) {
+	if (xgene_pmu->version == PCP_PMU_V3) {
+		intr_mcu = PCPPMU_V3_INT_MCU;
+		intr_mcb = PCPPMU_V3_INT_MCB;
+		intr_l3c = PCPPMU_V3_INT_L3C;
+		intr_iob = PCPPMU_V3_INT_IOB;
+	} else {
+		intr_mcu = PCPPMU_INT_MCU;
+		intr_mcb = PCPPMU_INT_MCB;
+		intr_l3c = PCPPMU_INT_L3C;
+		intr_iob = PCPPMU_INT_IOB;
+	}
+	if (val & intr_mcu) {
 		list_for_each_entry(ctx, &xgene_pmu->mcpmus, next) {
 			_xgene_pmu_isr(irq, ctx->pmu_dev);
 		}
 	}
-	if (val & PCPPMU_INT_MCB) {
+	if (val & intr_mcb) {
 		list_for_each_entry(ctx, &xgene_pmu->mcbpmus, next) {
 			_xgene_pmu_isr(irq, ctx->pmu_dev);
 		}
 	}
-	if (val & PCPPMU_INT_L3C) {
+	if (val & intr_l3c) {
 		list_for_each_entry(ctx, &xgene_pmu->l3cpmus, next) {
 			_xgene_pmu_isr(irq, ctx->pmu_dev);
 		}
 	}
-	if (val & PCPPMU_INT_IOB) {
+	if (val & intr_iob) {
 		list_for_each_entry(ctx, &xgene_pmu->iobpmus, next) {
 			_xgene_pmu_isr(irq, ctx->pmu_dev);
 		}
@@ -857,8 +1299,8 @@ static irqreturn_t xgene_pmu_isr(int irq, void *dev_id)
 	return IRQ_HANDLED;
 }
 
-static int acpi_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
-					 struct platform_device *pdev)
+static int acpi_pmu_probe_active_mcb_mcu_l3c(struct xgene_pmu *xgene_pmu,
+					     struct platform_device *pdev)
 {
 	void __iomem *csw_csr, *mcba_csr, *mcbb_csr;
 	struct resource *res;
@@ -885,6 +1327,8 @@ static int acpi_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
 		return PTR_ERR(mcbb_csr);
 	}
 
+	xgene_pmu->l3c_active_mask = 0x1;
+
 	reg = readl(csw_csr + CSW_CSWCR);
 	if (reg & CSW_CSWCR_DUALMCB_MASK) {
 		/* Dual MCB active */
@@ -905,8 +1349,56 @@ static int acpi_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
 	return 0;
 }
 
-static int fdt_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
-					struct platform_device *pdev)
+static int acpi_pmu_v3_probe_active_mcb_mcu_l3c(struct xgene_pmu *xgene_pmu,
+						struct platform_device *pdev)
+{
+	void __iomem *csw_csr;
+	struct resource *res;
+	unsigned int reg;
+	u32 mcb0routing;
+	u32 mcb1routing;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	csw_csr = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(csw_csr)) {
+		dev_err(&pdev->dev, "ioremap failed for CSW CSR resource\n");
+		return PTR_ERR(csw_csr);
+	}
+
+	reg = readl(csw_csr + CSW_CSWCR);
+	mcb0routing = CSW_CSWCR_MCB0_ROUTING(reg);
+	mcb1routing = CSW_CSWCR_MCB1_ROUTING(reg);
+	if (reg & CSW_CSWCR_DUALMCB_MASK) {
+		/* Dual MCB active */
+		xgene_pmu->mcb_active_mask = 0x3;
+		/* Probe all active L3C(s), maximum is 8 */
+		xgene_pmu->l3c_active_mask = 0xFF;
+		/* Probe all active MC(s), maximum is 8 */
+		if ((mcb0routing == 0x2) && (mcb1routing == 0x2))
+			xgene_pmu->mc_active_mask = 0xFF;
+		else if ((mcb0routing == 0x1) && (mcb1routing == 0x1))
+			xgene_pmu->mc_active_mask =  0x33;
+		else
+			xgene_pmu->mc_active_mask =  0x11;
+	} else {
+		/* Single MCB active */
+		xgene_pmu->mcb_active_mask = 0x1;
+		/* Probe all active L3C(s), maximum is 4 */
+		xgene_pmu->l3c_active_mask = 0x0F;
+		/* Probe all active MC(s), maximum is 4 */
+		if (mcb0routing == 0x2)
+			xgene_pmu->mc_active_mask = 0x0F;
+		else if (mcb0routing == 0x1)
+			xgene_pmu->mc_active_mask =  0x03;
+		else
+			xgene_pmu->mc_active_mask =  0x01;
+	}
+
+	return 0;
+}
+
+static int fdt_pmu_probe_active_mcb_mcu_l3c(struct xgene_pmu *xgene_pmu,
+					    struct platform_device *pdev)
 {
 	struct regmap *csw_map, *mcba_map, *mcbb_map;
 	struct device_node *np = pdev->dev.of_node;
@@ -930,6 +1422,7 @@ static int fdt_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
 		return PTR_ERR(mcbb_map);
 	}
 
+	xgene_pmu->l3c_active_mask = 0x1;
 	if (regmap_read(csw_map, CSW_CSWCR, &reg))
 		return -EINVAL;
 
@@ -954,12 +1447,18 @@ static int fdt_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
 	return 0;
 }
 
-static int xgene_pmu_probe_active_mcb_mcu(struct xgene_pmu *xgene_pmu,
-					  struct platform_device *pdev)
+static int xgene_pmu_probe_active_mcb_mcu_l3c(struct xgene_pmu *xgene_pmu,
+					      struct platform_device *pdev)
 {
-	if (has_acpi_companion(&pdev->dev))
-		return acpi_pmu_probe_active_mcb_mcu(xgene_pmu, pdev);
-	return fdt_pmu_probe_active_mcb_mcu(xgene_pmu, pdev);
+	if (has_acpi_companion(&pdev->dev)) {
+		if (xgene_pmu->version == PCP_PMU_V3)
+			return acpi_pmu_v3_probe_active_mcb_mcu_l3c(xgene_pmu,
+								    pdev);
+		else
+			return acpi_pmu_probe_active_mcb_mcu_l3c(xgene_pmu,
+								 pdev);
+	}
+	return fdt_pmu_probe_active_mcb_mcu_l3c(xgene_pmu, pdev);
 }
 
 static char *xgene_pmu_dev_name(struct device *dev, u32 type, int id)
@@ -969,6 +1468,8 @@ static char *xgene_pmu_dev_name(struct device *dev, u32 type, int id)
 		return devm_kasprintf(dev, GFP_KERNEL, "l3c%d", id);
 	case PMU_TYPE_IOB:
 		return devm_kasprintf(dev, GFP_KERNEL, "iob%d", id);
+	case PMU_TYPE_IOB_SLOW:
+		return devm_kasprintf(dev, GFP_KERNEL, "iob-slow%d", id);
 	case PMU_TYPE_MCB:
 		return devm_kasprintf(dev, GFP_KERNEL, "mcb%d", id);
 	case PMU_TYPE_MC:
@@ -1047,9 +1548,40 @@ err:
 	return NULL;
 }
 
+static const struct acpi_device_id xgene_pmu_acpi_type_match[] = {
+	{"APMC0D5D", PMU_TYPE_L3C},
+	{"APMC0D5E", PMU_TYPE_IOB},
+	{"APMC0D5F", PMU_TYPE_MCB},
+	{"APMC0D60", PMU_TYPE_MC},
+	{"APMC0D84", PMU_TYPE_L3C},
+	{"APMC0D85", PMU_TYPE_IOB},
+	{"APMC0D86", PMU_TYPE_IOB_SLOW},
+	{"APMC0D87", PMU_TYPE_MCB},
+	{"APMC0D88", PMU_TYPE_MC},
+	{},
+};
+
+static const struct acpi_device_id *xgene_pmu_acpi_match_type(
+					const struct acpi_device_id *ids,
+					struct acpi_device *adev)
+{
+	const struct acpi_device_id *match_id = NULL;
+	const struct acpi_device_id *id;
+
+	for (id = ids; id->id[0] || id->cls; id++) {
+		if (!acpi_match_device_ids(adev, id))
+			match_id = id;
+		else if (match_id)
+			break;
+	}
+
+	return match_id;
+}
+
 static acpi_status acpi_pmu_dev_add(acpi_handle handle, u32 level,
 				    void *data, void **return_value)
 {
+	const struct acpi_device_id *acpi_id;
 	struct xgene_pmu *xgene_pmu = data;
 	struct xgene_pmu_dev_ctx *ctx;
 	struct acpi_device *adev;
@@ -1059,17 +1591,11 @@ static acpi_status acpi_pmu_dev_add(acpi_handle handle, u32 level,
 	if (acpi_bus_get_status(adev) || !adev->status.present)
 		return AE_OK;
 
-	if (!strcmp(acpi_device_hid(adev), "APMC0D5D"))
-		ctx = acpi_get_pmu_hw_inf(xgene_pmu, adev, PMU_TYPE_L3C);
-	else if (!strcmp(acpi_device_hid(adev), "APMC0D5E"))
-		ctx = acpi_get_pmu_hw_inf(xgene_pmu, adev, PMU_TYPE_IOB);
-	else if (!strcmp(acpi_device_hid(adev), "APMC0D5F"))
-		ctx = acpi_get_pmu_hw_inf(xgene_pmu, adev, PMU_TYPE_MCB);
-	else if (!strcmp(acpi_device_hid(adev), "APMC0D60"))
-		ctx = acpi_get_pmu_hw_inf(xgene_pmu, adev, PMU_TYPE_MC);
-	else
-		ctx = NULL;
+	acpi_id = xgene_pmu_acpi_match_type(xgene_pmu_acpi_type_match, adev);
+	if (!acpi_id)
+		return AE_OK;
 
+	ctx = acpi_get_pmu_hw_inf(xgene_pmu, adev, (u32)acpi_id->driver_data);
 	if (!ctx)
 		return AE_OK;
 
@@ -1086,6 +1612,9 @@ static acpi_status acpi_pmu_dev_add(acpi_handle handle, u32 level,
 	case PMU_TYPE_IOB:
 		list_add(&ctx->next, &xgene_pmu->iobpmus);
 		break;
+	case PMU_TYPE_IOB_SLOW:
+		list_add(&ctx->next, &xgene_pmu->iobpmus);
+		break;
 	case PMU_TYPE_MCB:
 		list_add(&ctx->next, &xgene_pmu->mcbpmus);
 		break;
@@ -1207,6 +1736,9 @@ static int fdt_pmu_probe_pmu_dev(struct xgene_pmu *xgene_pmu,
 		case PMU_TYPE_IOB:
 			list_add(&ctx->next, &xgene_pmu->iobpmus);
 			break;
+		case PMU_TYPE_IOB_SLOW:
+			list_add(&ctx->next, &xgene_pmu->iobpmus);
+			break;
 		case PMU_TYPE_MCB:
 			list_add(&ctx->next, &xgene_pmu->mcbpmus);
 			break;
@@ -1235,6 +1767,40 @@ static const struct xgene_pmu_data xgene_pmu_v2_data = {
 	.id   = PCP_PMU_V2,
 };
 
+static const struct xgene_pmu_ops xgene_pmu_ops = {
+	.mask_int = xgene_pmu_mask_int,
+	.unmask_int = xgene_pmu_unmask_int,
+	.read_counter = xgene_pmu_read_counter32,
+	.write_counter = xgene_pmu_write_counter32,
+	.write_evttype = xgene_pmu_write_evttype,
+	.write_agentmsk = xgene_pmu_write_agentmsk,
+	.write_agent1msk = xgene_pmu_write_agent1msk,
+	.enable_counter = xgene_pmu_enable_counter,
+	.disable_counter = xgene_pmu_disable_counter,
+	.enable_counter_int = xgene_pmu_enable_counter_int,
+	.disable_counter_int = xgene_pmu_disable_counter_int,
+	.reset_counters = xgene_pmu_reset_counters,
+	.start_counters = xgene_pmu_start_counters,
+	.stop_counters = xgene_pmu_stop_counters,
+};
+
+static const struct xgene_pmu_ops xgene_pmu_v3_ops = {
+	.mask_int = xgene_pmu_v3_mask_int,
+	.unmask_int = xgene_pmu_v3_unmask_int,
+	.read_counter = xgene_pmu_read_counter64,
+	.write_counter = xgene_pmu_write_counter64,
+	.write_evttype = xgene_pmu_write_evttype,
+	.write_agentmsk = xgene_pmu_v3_write_agentmsk,
+	.write_agent1msk = xgene_pmu_v3_write_agent1msk,
+	.enable_counter = xgene_pmu_enable_counter,
+	.disable_counter = xgene_pmu_disable_counter,
+	.enable_counter_int = xgene_pmu_enable_counter_int,
+	.disable_counter_int = xgene_pmu_disable_counter_int,
+	.reset_counters = xgene_pmu_reset_counters,
+	.start_counters = xgene_pmu_start_counters,
+	.stop_counters = xgene_pmu_stop_counters,
+};
+
 static const struct of_device_id xgene_pmu_of_match[] = {
 	{ .compatible	= "apm,xgene-pmu",	.data = &xgene_pmu_data },
 	{ .compatible	= "apm,xgene-pmu-v2",	.data = &xgene_pmu_v2_data },
@@ -1245,6 +1811,7 @@ MODULE_DEVICE_TABLE(of, xgene_pmu_of_match);
 static const struct acpi_device_id xgene_pmu_acpi_match[] = {
 	{"APMC0D5B", PCP_PMU_V1},
 	{"APMC0D5C", PCP_PMU_V2},
+	{"APMC0D83", PCP_PMU_V3},
 	{},
 };
 MODULE_DEVICE_TABLE(acpi, xgene_pmu_acpi_match);
@@ -1284,6 +1851,11 @@ static int xgene_pmu_probe(struct platform_device *pdev)
 	if (version < 0)
 		return -ENODEV;
 
+	if (version == PCP_PMU_V3)
+		xgene_pmu->ops = &xgene_pmu_v3_ops;
+	else
+		xgene_pmu->ops = &xgene_pmu_ops;
+
 	INIT_LIST_HEAD(&xgene_pmu->l3cpmus);
 	INIT_LIST_HEAD(&xgene_pmu->iobpmus);
 	INIT_LIST_HEAD(&xgene_pmu->mcbpmus);
@@ -1317,7 +1889,7 @@ static int xgene_pmu_probe(struct platform_device *pdev)
 	raw_spin_lock_init(&xgene_pmu->lock);
 
 	/* Check for active MCBs and MCUs */
-	rc = xgene_pmu_probe_active_mcb_mcu(xgene_pmu, pdev);
+	rc = xgene_pmu_probe_active_mcb_mcu_l3c(xgene_pmu, pdev);
 	if (rc) {
 		dev_warn(&pdev->dev, "Unknown MCB/MCU active status\n");
 		xgene_pmu->mcb_active_mask = 0x1;
@@ -1342,7 +1914,7 @@ static int xgene_pmu_probe(struct platform_device *pdev)
 	}
 
 	/* Enable interrupt */
-	xgene_pmu_unmask_int(xgene_pmu);
+	xgene_pmu->ops->unmask_int(xgene_pmu);
 
 	return 0;
 
diff --git a/drivers/ras/ras.c b/drivers/ras/ras.c
index ed4c343d08c4..5429d3795732 100644
--- a/drivers/ras/ras.c
+++ b/drivers/ras/ras.c
@@ -7,11 +7,24 @@
 
 #include <linux/init.h>
 #include <linux/ras.h>
+#include <linux/uuid.h>
 
 #define CREATE_TRACE_POINTS
 #define TRACE_INCLUDE_PATH ../../include/ras
 #include <ras/ras_event.h>
 
+void log_non_standard_event(const uuid_le *sec_type, const uuid_le *fru_id,
+			    const char *fru_text, const u8 sev, const u8 *err,
+			    const u32 len)
+{
+	trace_non_standard_event(sec_type, fru_id, fru_text, sev, err, len);
+}
+
+void log_arm_hw_error(struct cper_sec_proc_arm *err)
+{
+	trace_arm_event(err);
+}
+
 static int __init ras_init(void)
 {
 	int rc = 0;
@@ -27,7 +40,8 @@ subsys_initcall(ras_init);
 EXPORT_TRACEPOINT_SYMBOL_GPL(extlog_mem_event);
 #endif
 EXPORT_TRACEPOINT_SYMBOL_GPL(mc_event);
-
+EXPORT_TRACEPOINT_SYMBOL_GPL(non_standard_event);
+EXPORT_TRACEPOINT_SYMBOL_GPL(arm_event);
 
 static int __init parse_ras_param(char *str)
 {