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

Pull arm64 updates from Catalin Marinas:

 - Expose tag address bits in siginfo. The original arm64 ABI did not
   expose any of the bits 63:56 of a tagged address in siginfo. In the
   presence of user ASAN or MTE, this information may be useful. The
   implementation is generic to other architectures supporting tags
   (like SPARC ADI, subject to wiring up the arch code). The user will
   have to opt in via sigaction(SA_EXPOSE_TAGBITS) so that the extra
   bits, if available, become visible in si_addr.

 - Default to 32-bit wide ZONE_DMA. Previously, ZONE_DMA was set to the
   lowest 1GB to cope with the Raspberry Pi 4 limitations, to the
   detriment of other platforms. With these changes, the kernel scans
   the Device Tree dma-ranges and the ACPI IORT information before
   deciding on a smaller ZONE_DMA.

 - Strengthen READ_ONCE() to acquire when CONFIG_LTO=y. When building
   with LTO, there is an increased risk of the compiler converting an
   address dependency headed by a READ_ONCE() invocation into a control
   dependency and consequently allowing for harmful reordering by the
   CPU.

 - Add CPPC FFH support using arm64 AMU counters.

 - set_fs() removal on arm64. This renders the User Access Override
   (UAO) ARMv8 feature unnecessary.

 - Perf updates: PMU driver for the ARM DMC-620 memory controller, sysfs
   identifier file for SMMUv3, stop event counters support for i.MX8MP,
   enable the perf events-based hard lockup detector.

 - Reorganise the kernel VA space slightly so that 52-bit VA
   configurations can use more virtual address space.

 - Improve the robustness of the arm64 memory offline event notifier.

 - Pad the Image header to 64K following the EFI header definition
   updated recently to increase the section alignment to 64K.

 - Support CONFIG_CMDLINE_EXTEND on arm64.

 - Do not use tagged PC in the kernel (TCR_EL1.TBID1==1), freeing up 8
   bits for PtrAuth.

 - Switch to vmapped shadow call stacks.

 - Miscellaneous clean-ups.

* tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux: (78 commits)
  perf/imx_ddr: Add system PMU identifier for userspace
  bindings: perf: imx-ddr: add compatible string
  arm64: Fix build failure when HARDLOCKUP_DETECTOR_PERF is enabled
  arm64: mte: fix prctl(PR_GET_TAGGED_ADDR_CTRL) if TCF0=NONE
  arm64: mark __system_matches_cap as __maybe_unused
  arm64: uaccess: remove vestigal UAO support
  arm64: uaccess: remove redundant PAN toggling
  arm64: uaccess: remove addr_limit_user_check()
  arm64: uaccess: remove set_fs()
  arm64: uaccess cleanup macro naming
  arm64: uaccess: split user/kernel routines
  arm64: uaccess: refactor __{get,put}_user
  arm64: uaccess: simplify __copy_user_flushcache()
  arm64: uaccess: rename privileged uaccess routines
  arm64: sdei: explicitly simulate PAN/UAO entry
  arm64: sdei: move uaccess logic to arch/arm64/
  arm64: head.S: always initialize PSTATE
  arm64: head.S: cleanup SCTLR_ELx initialization
  arm64: head.S: rename el2_setup -> init_kernel_el
  arm64: add C wrappers for SET_PSTATE_*()
  ...

16 files changed, 1092 insertions, 43 deletions

diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c
index 770d84071a32..d4eac6d7e9fb 100644
--- a/drivers/acpi/arm64/iort.c
+++ b/drivers/acpi/arm64/iort.c
@@ -1720,3 +1720,58 @@ void __init acpi_iort_init(void)
 
 	iort_init_platform_devices();
 }
+
+#ifdef CONFIG_ZONE_DMA
+/*
+ * Extract the highest CPU physical address accessible to all DMA masters in
+ * the system. PHYS_ADDR_MAX is returned when no constrained device is found.
+ */
+phys_addr_t __init acpi_iort_dma_get_max_cpu_address(void)
+{
+	phys_addr_t limit = PHYS_ADDR_MAX;
+	struct acpi_iort_node *node, *end;
+	struct acpi_table_iort *iort;
+	acpi_status status;
+	int i;
+
+	if (acpi_disabled)
+		return limit;
+
+	status = acpi_get_table(ACPI_SIG_IORT, 0,
+				(struct acpi_table_header **)&iort);
+	if (ACPI_FAILURE(status))
+		return limit;
+
+	node = ACPI_ADD_PTR(struct acpi_iort_node, iort, iort->node_offset);
+	end = ACPI_ADD_PTR(struct acpi_iort_node, iort, iort->header.length);
+
+	for (i = 0; i < iort->node_count; i++) {
+		if (node >= end)
+			break;
+
+		switch (node->type) {
+			struct acpi_iort_named_component *ncomp;
+			struct acpi_iort_root_complex *rc;
+			phys_addr_t local_limit;
+
+		case ACPI_IORT_NODE_NAMED_COMPONENT:
+			ncomp = (struct acpi_iort_named_component *)node->node_data;
+			local_limit = DMA_BIT_MASK(ncomp->memory_address_limit);
+			limit = min_not_zero(limit, local_limit);
+			break;
+
+		case ACPI_IORT_NODE_PCI_ROOT_COMPLEX:
+			if (node->revision < 1)
+				break;
+
+			rc = (struct acpi_iort_root_complex *)node->node_data;
+			local_limit = DMA_BIT_MASK(rc->memory_address_limit);
+			limit = min_not_zero(limit, local_limit);
+			break;
+		}
+		node = ACPI_ADD_PTR(struct acpi_iort_node, node, node->length);
+	}
+	acpi_put_table(&iort->header);
+	return limit;
+}
+#endif
diff --git a/drivers/firmware/arm_sdei.c b/drivers/firmware/arm_sdei.c
index 840754dcc6ca..a7e762c352f9 100644
--- a/drivers/firmware/arm_sdei.c
+++ b/drivers/firmware/arm_sdei.c
@@ -31,7 +31,6 @@
 #include <linux/slab.h>
 #include <linux/smp.h>
 #include <linux/spinlock.h>
-#include <linux/uaccess.h>
 
 /*
  * The call to use to reach the firmware.
@@ -1092,26 +1091,13 @@ int sdei_event_handler(struct pt_regs *regs,
 		       struct sdei_registered_event *arg)
 {
 	int err;
-	mm_segment_t orig_addr_limit;
 	u32 event_num = arg->event_num;
 
-	/*
-	 * Save restore 'fs'.
-	 * The architecture's entry code save/restores 'fs' when taking an
-	 * exception from the kernel. This ensures addr_limit isn't inherited
-	 * if you interrupted something that allowed the uaccess routines to
-	 * access kernel memory.
-	 * Do the same here because this doesn't come via the same entry code.
-	*/
-	orig_addr_limit = force_uaccess_begin();
-
 	err = arg->callback(event_num, regs, arg->callback_arg);
 	if (err)
 		pr_err_ratelimited("event %u on CPU %u failed with error: %d\n",
 				   event_num, smp_processor_id(), err);
 
-	force_uaccess_end(orig_addr_limit);
-
 	return err;
 }
 NOKPROBE_SYMBOL(sdei_event_handler);
diff --git a/drivers/of/address.c b/drivers/of/address.c
index 1c3257a2d4e3..73ddf2540f3f 100644
--- a/drivers/of/address.c
+++ b/drivers/of/address.c
@@ -1025,6 +1025,48 @@ out:
 #endif /* CONFIG_HAS_DMA */
 
 /**
+ * of_dma_get_max_cpu_address - Gets highest CPU address suitable for DMA
+ * @np: The node to start searching from or NULL to start from the root
+ *
+ * Gets the highest CPU physical address that is addressable by all DMA masters
+ * in the sub-tree pointed by np, or the whole tree if NULL is passed. If no
+ * DMA constrained device is found, it returns PHYS_ADDR_MAX.
+ */
+phys_addr_t __init of_dma_get_max_cpu_address(struct device_node *np)
+{
+	phys_addr_t max_cpu_addr = PHYS_ADDR_MAX;
+	struct of_range_parser parser;
+	phys_addr_t subtree_max_addr;
+	struct device_node *child;
+	struct of_range range;
+	const __be32 *ranges;
+	u64 cpu_end = 0;
+	int len;
+
+	if (!np)
+		np = of_root;
+
+	ranges = of_get_property(np, "dma-ranges", &len);
+	if (ranges && len) {
+		of_dma_range_parser_init(&parser, np);
+		for_each_of_range(&parser, &range)
+			if (range.cpu_addr + range.size > cpu_end)
+				cpu_end = range.cpu_addr + range.size - 1;
+
+		if (max_cpu_addr > cpu_end)
+			max_cpu_addr = cpu_end;
+	}
+
+	for_each_available_child_of_node(np, child) {
+		subtree_max_addr = of_dma_get_max_cpu_address(child);
+		if (max_cpu_addr > subtree_max_addr)
+			max_cpu_addr = subtree_max_addr;
+	}
+
+	return max_cpu_addr;
+}
+
+/**
  * of_dma_is_coherent - Check if device is coherent
  * @np:	device node
  *
diff --git a/drivers/of/unittest.c b/drivers/of/unittest.c
index 06cc988faf78..eb51bc147440 100644
--- a/drivers/of/unittest.c
+++ b/drivers/of/unittest.c
@@ -869,6 +869,26 @@ static void __init of_unittest_changeset(void)
 #endif
 }
 
+static void __init of_unittest_dma_get_max_cpu_address(void)
+{
+	struct device_node *np;
+	phys_addr_t cpu_addr;
+
+	if (!IS_ENABLED(CONFIG_OF_ADDRESS))
+		return;
+
+	np = of_find_node_by_path("/testcase-data/address-tests");
+	if (!np) {
+		pr_err("missing testcase data\n");
+		return;
+	}
+
+	cpu_addr = of_dma_get_max_cpu_address(np);
+	unittest(cpu_addr == 0x4fffffff,
+		 "of_dma_get_max_cpu_address: wrong CPU addr %pad (expecting %x)\n",
+		 &cpu_addr, 0x4fffffff);
+}
+
 static void __init of_unittest_dma_ranges_one(const char *path,
 		u64 expect_dma_addr, u64 expect_paddr)
 {
@@ -3266,6 +3286,7 @@ static int __init of_unittest(void)
 	of_unittest_changeset();
 	of_unittest_parse_interrupts();
 	of_unittest_parse_interrupts_extended();
+	of_unittest_dma_get_max_cpu_address();
 	of_unittest_parse_dma_ranges();
 	of_unittest_pci_dma_ranges();
 	of_unittest_match_node();
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index 130327ff0b0e..3075cf171f78 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -130,6 +130,13 @@ config ARM_SPE_PMU
 	  Extension, which provides periodic sampling of operations in
 	  the CPU pipeline and reports this via the perf AUX interface.
 
+config ARM_DMC620_PMU
+	tristate "Enable PMU support for the ARM DMC-620 memory controller"
+	depends on (ARM64 && ACPI) || COMPILE_TEST
+	help
+	  Support for PMU events monitoring on the ARM DMC-620 memory
+	  controller.
+
 source "drivers/perf/hisilicon/Kconfig"
 
 endmenu
diff --git a/drivers/perf/Makefile b/drivers/perf/Makefile
index 5365fd56f88f..5260b116c7da 100644
--- a/drivers/perf/Makefile
+++ b/drivers/perf/Makefile
@@ -13,3 +13,4 @@ obj-$(CONFIG_QCOM_L3_PMU) += qcom_l3_pmu.o
 obj-$(CONFIG_THUNDERX2_PMU) += thunderx2_pmu.o
 obj-$(CONFIG_XGENE_PMU) += xgene_pmu.o
 obj-$(CONFIG_ARM_SPE_PMU) += arm_spe_pmu.o
+obj-$(CONFIG_ARM_DMC620_PMU) += arm_dmc620_pmu.o
diff --git a/drivers/perf/arm_dmc620_pmu.c b/drivers/perf/arm_dmc620_pmu.c
new file mode 100644
index 000000000000..004930eb4bbb
--- /dev/null
+++ b/drivers/perf/arm_dmc620_pmu.c
@@ -0,0 +1,748 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * ARM DMC-620 memory controller PMU driver
+ *
+ * Copyright (C) 2020 Ampere Computing LLC.
+ */
+
+#define DMC620_PMUNAME		"arm_dmc620"
+#define DMC620_DRVNAME		DMC620_PMUNAME "_pmu"
+#define pr_fmt(fmt)		DMC620_DRVNAME ": " fmt
+
+#include <linux/acpi.h>
+#include <linux/bitfield.h>
+#include <linux/bitops.h>
+#include <linux/cpuhotplug.h>
+#include <linux/cpumask.h>
+#include <linux/device.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/perf_event.h>
+#include <linux/platform_device.h>
+#include <linux/printk.h>
+#include <linux/rculist.h>
+#include <linux/refcount.h>
+
+#define DMC620_PA_SHIFT					12
+#define DMC620_CNT_INIT					0x80000000
+#define DMC620_CNT_MAX_PERIOD				0xffffffff
+#define DMC620_PMU_CLKDIV2_MAX_COUNTERS			8
+#define DMC620_PMU_CLK_MAX_COUNTERS			2
+#define DMC620_PMU_MAX_COUNTERS				\
+	(DMC620_PMU_CLKDIV2_MAX_COUNTERS + DMC620_PMU_CLK_MAX_COUNTERS)
+
+/*
+ * The PMU registers start at 0xA00 in the DMC-620 memory map, and these
+ * offsets are relative to that base.
+ *
+ * Each counter has a group of control/value registers, and the
+ * DMC620_PMU_COUNTERn offsets are within a counter group.
+ *
+ * The counter registers groups start at 0xA10.
+ */
+#define DMC620_PMU_OVERFLOW_STATUS_CLKDIV2		0x8
+#define  DMC620_PMU_OVERFLOW_STATUS_CLKDIV2_MASK	\
+		(DMC620_PMU_CLKDIV2_MAX_COUNTERS - 1)
+#define DMC620_PMU_OVERFLOW_STATUS_CLK			0xC
+#define  DMC620_PMU_OVERFLOW_STATUS_CLK_MASK		\
+		(DMC620_PMU_CLK_MAX_COUNTERS - 1)
+#define DMC620_PMU_COUNTERS_BASE			0x10
+#define DMC620_PMU_COUNTERn_MASK_31_00			0x0
+#define DMC620_PMU_COUNTERn_MASK_63_32			0x4
+#define DMC620_PMU_COUNTERn_MATCH_31_00			0x8
+#define DMC620_PMU_COUNTERn_MATCH_63_32			0xC
+#define DMC620_PMU_COUNTERn_CONTROL			0x10
+#define  DMC620_PMU_COUNTERn_CONTROL_ENABLE		BIT(0)
+#define  DMC620_PMU_COUNTERn_CONTROL_INVERT		BIT(1)
+#define  DMC620_PMU_COUNTERn_CONTROL_EVENT_MUX		GENMASK(6, 2)
+#define  DMC620_PMU_COUNTERn_CONTROL_INCR_MUX		GENMASK(8, 7)
+#define DMC620_PMU_COUNTERn_VALUE			0x20
+/* Offset of the registers for a given counter, relative to 0xA00 */
+#define DMC620_PMU_COUNTERn_OFFSET(n) \
+	(DMC620_PMU_COUNTERS_BASE + 0x28 * (n))
+
+static LIST_HEAD(dmc620_pmu_irqs);
+static DEFINE_MUTEX(dmc620_pmu_irqs_lock);
+
+struct dmc620_pmu_irq {
+	struct hlist_node node;
+	struct list_head pmus_node;
+	struct list_head irqs_node;
+	refcount_t refcount;
+	unsigned int irq_num;
+	unsigned int cpu;
+};
+
+struct dmc620_pmu {
+	struct pmu pmu;
+
+	void __iomem *base;
+	struct dmc620_pmu_irq *irq;
+	struct list_head pmus_node;
+
+	/*
+	 * We put all clkdiv2 and clk counters to a same array.
+	 * The first DMC620_PMU_CLKDIV2_MAX_COUNTERS bits belong to
+	 * clkdiv2 counters, the last DMC620_PMU_CLK_MAX_COUNTERS
+	 * belong to clk counters.
+	 */
+	DECLARE_BITMAP(used_mask, DMC620_PMU_MAX_COUNTERS);
+	struct perf_event *events[DMC620_PMU_MAX_COUNTERS];
+};
+
+#define to_dmc620_pmu(p) (container_of(p, struct dmc620_pmu, pmu))
+
+static int cpuhp_state_num;
+
+struct dmc620_pmu_event_attr {
+	struct device_attribute attr;
+	u8 clkdiv2;
+	u8 eventid;
+};
+
+static ssize_t
+dmc620_pmu_event_show(struct device *dev,
+			   struct device_attribute *attr, char *page)
+{
+	struct dmc620_pmu_event_attr *eattr;
+
+	eattr = container_of(attr, typeof(*eattr), attr);
+
+	return sprintf(page, "event=0x%x,clkdiv2=0x%x\n", eattr->eventid, eattr->clkdiv2);
+}
+
+#define DMC620_PMU_EVENT_ATTR(_name, _eventid, _clkdiv2)		\
+	(&((struct dmc620_pmu_event_attr[]) {{				\
+		.attr = __ATTR(_name, 0444, dmc620_pmu_event_show, NULL),	\
+		.clkdiv2 = _clkdiv2,						\
+		.eventid = _eventid,					\
+	}})[0].attr.attr)
+
+static struct attribute *dmc620_pmu_events_attrs[] = {
+	/* clkdiv2 events list */
+	DMC620_PMU_EVENT_ATTR(clkdiv2_cycle_count, 0x0, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_allocate, 0x1, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_queue_depth, 0x2, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_waiting_for_wr_data, 0x3, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_read_backlog, 0x4, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_waiting_for_mi, 0x5, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_hazard_resolution, 0x6, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_enqueue, 0x7, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_arbitrate, 0x8, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_lrank_turnaround_activate, 0x9, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_prank_turnaround_activate, 0xa, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_read_depth, 0xb, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_write_depth, 0xc, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_highigh_qos_depth, 0xd, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_high_qos_depth, 0xe, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_medium_qos_depth, 0xf, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_low_qos_depth, 0x10, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_activate, 0x11, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_rdwr, 0x12, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_refresh, 0x13, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_training_request, 0x14, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_t_mac_tracker, 0x15, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_bk_fsm_tracker, 0x16, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_bk_open_tracker, 0x17, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_ranks_in_pwr_down, 0x18, 1),
+	DMC620_PMU_EVENT_ATTR(clkdiv2_ranks_in_sref, 0x19, 1),
+
+	/* clk events list */
+	DMC620_PMU_EVENT_ATTR(clk_cycle_count, 0x0, 0),
+	DMC620_PMU_EVENT_ATTR(clk_request, 0x1, 0),
+	DMC620_PMU_EVENT_ATTR(clk_upload_stall, 0x2, 0),
+	NULL,
+};
+
+static struct attribute_group dmc620_pmu_events_attr_group = {
+	.name = "events",
+	.attrs = dmc620_pmu_events_attrs,
+};
+
+/* User ABI */
+#define ATTR_CFG_FLD_mask_CFG		config
+#define ATTR_CFG_FLD_mask_LO		0
+#define ATTR_CFG_FLD_mask_HI		44
+#define ATTR_CFG_FLD_match_CFG		config1
+#define ATTR_CFG_FLD_match_LO		0
+#define ATTR_CFG_FLD_match_HI		44
+#define ATTR_CFG_FLD_invert_CFG		config2
+#define ATTR_CFG_FLD_invert_LO		0
+#define ATTR_CFG_FLD_invert_HI		0
+#define ATTR_CFG_FLD_incr_CFG		config2
+#define ATTR_CFG_FLD_incr_LO		1
+#define ATTR_CFG_FLD_incr_HI		2
+#define ATTR_CFG_FLD_event_CFG		config2
+#define ATTR_CFG_FLD_event_LO		3
+#define ATTR_CFG_FLD_event_HI		8
+#define ATTR_CFG_FLD_clkdiv2_CFG	config2
+#define ATTR_CFG_FLD_clkdiv2_LO		9
+#define ATTR_CFG_FLD_clkdiv2_HI		9
+
+#define __GEN_PMU_FORMAT_ATTR(cfg, lo, hi)			\
+	(lo) == (hi) ? #cfg ":" #lo "\n" : #cfg ":" #lo "-" #hi
+
+#define _GEN_PMU_FORMAT_ATTR(cfg, lo, hi)			\
+	__GEN_PMU_FORMAT_ATTR(cfg, lo, hi)
+
+#define GEN_PMU_FORMAT_ATTR(name)				\
+	PMU_FORMAT_ATTR(name,					\
+	_GEN_PMU_FORMAT_ATTR(ATTR_CFG_FLD_##name##_CFG,		\
+			     ATTR_CFG_FLD_##name##_LO,		\
+			     ATTR_CFG_FLD_##name##_HI))
+
+#define _ATTR_CFG_GET_FLD(attr, cfg, lo, hi)			\
+	((((attr)->cfg) >> lo) & GENMASK_ULL(hi - lo, 0))
+
+#define ATTR_CFG_GET_FLD(attr, name)				\
+	_ATTR_CFG_GET_FLD(attr,					\
+			  ATTR_CFG_FLD_##name##_CFG,		\
+			  ATTR_CFG_FLD_##name##_LO,		\
+			  ATTR_CFG_FLD_##name##_HI)
+
+GEN_PMU_FORMAT_ATTR(mask);
+GEN_PMU_FORMAT_ATTR(match);
+GEN_PMU_FORMAT_ATTR(invert);
+GEN_PMU_FORMAT_ATTR(incr);
+GEN_PMU_FORMAT_ATTR(event);
+GEN_PMU_FORMAT_ATTR(clkdiv2);
+
+static struct attribute *dmc620_pmu_formats_attrs[] = {
+	&format_attr_mask.attr,
+	&format_attr_match.attr,
+	&format_attr_invert.attr,
+	&format_attr_incr.attr,
+	&format_attr_event.attr,
+	&format_attr_clkdiv2.attr,
+	NULL,
+};
+
+static struct attribute_group dmc620_pmu_format_attr_group = {
+	.name	= "format",
+	.attrs	= dmc620_pmu_formats_attrs,
+};
+
+static const struct attribute_group *dmc620_pmu_attr_groups[] = {
+	&dmc620_pmu_events_attr_group,
+	&dmc620_pmu_format_attr_group,
+	NULL,
+};
+
+static inline
+u32 dmc620_pmu_creg_read(struct dmc620_pmu *dmc620_pmu,
+			unsigned int idx, unsigned int reg)
+{
+	return readl(dmc620_pmu->base + DMC620_PMU_COUNTERn_OFFSET(idx) + reg);
+}
+
+static inline
+void dmc620_pmu_creg_write(struct dmc620_pmu *dmc620_pmu,
+			unsigned int idx, unsigned int reg, u32 val)
+{
+	writel(val, dmc620_pmu->base + DMC620_PMU_COUNTERn_OFFSET(idx) + reg);
+}
+
+static
+unsigned int dmc620_event_to_counter_control(struct perf_event *event)
+{
+	struct perf_event_attr *attr = &event->attr;
+	unsigned int reg = 0;
+
+	reg |= FIELD_PREP(DMC620_PMU_COUNTERn_CONTROL_INVERT,
+			ATTR_CFG_GET_FLD(attr, invert));
+	reg |= FIELD_PREP(DMC620_PMU_COUNTERn_CONTROL_EVENT_MUX,
+			ATTR_CFG_GET_FLD(attr, event));
+	reg |= FIELD_PREP(DMC620_PMU_COUNTERn_CONTROL_INCR_MUX,
+			ATTR_CFG_GET_FLD(attr, incr));
+
+	return reg;
+}
+
+static int dmc620_get_event_idx(struct perf_event *event)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+	int idx, start_idx, end_idx;
+
+	if (ATTR_CFG_GET_FLD(&event->attr, clkdiv2)) {
+		start_idx = 0;
+		end_idx = DMC620_PMU_CLKDIV2_MAX_COUNTERS;
+	} else {
+		start_idx = DMC620_PMU_CLKDIV2_MAX_COUNTERS;
+		end_idx = DMC620_PMU_MAX_COUNTERS;
+	}
+
+	for (idx = start_idx; idx < end_idx; ++idx) {
+		if (!test_and_set_bit(idx, dmc620_pmu->used_mask))
+			return idx;
+	}
+
+	/* The counters are all in use. */
+	return -EAGAIN;
+}
+
+static inline
+u64 dmc620_pmu_read_counter(struct perf_event *event)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+
+	return dmc620_pmu_creg_read(dmc620_pmu,
+				    event->hw.idx, DMC620_PMU_COUNTERn_VALUE);
+}
+
+static void dmc620_pmu_event_update(struct perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	u64 delta, prev_count, new_count;
+
+	do {
+		/* We may also be called from the irq handler */
+		prev_count = local64_read(&hwc->prev_count);
+		new_count = dmc620_pmu_read_counter(event);
+	} while (local64_cmpxchg(&hwc->prev_count,
+			prev_count, new_count) != prev_count);
+	delta = (new_count - prev_count) & DMC620_CNT_MAX_PERIOD;
+	local64_add(delta, &event->count);
+}
+
+static void dmc620_pmu_event_set_period(struct perf_event *event)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+
+	local64_set(&event->hw.prev_count, DMC620_CNT_INIT);
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      event->hw.idx, DMC620_PMU_COUNTERn_VALUE, DMC620_CNT_INIT);
+}
+
+static void dmc620_pmu_enable_counter(struct perf_event *event)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+	u32 reg;
+
+	reg = dmc620_event_to_counter_control(event) | DMC620_PMU_COUNTERn_CONTROL_ENABLE;
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      event->hw.idx, DMC620_PMU_COUNTERn_CONTROL, reg);
+}
+
+static void dmc620_pmu_disable_counter(struct perf_event *event)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      event->hw.idx, DMC620_PMU_COUNTERn_CONTROL, 0);
+}
+
+static irqreturn_t dmc620_pmu_handle_irq(int irq_num, void *data)
+{
+	struct dmc620_pmu_irq *irq = data;
+	struct dmc620_pmu *dmc620_pmu;
+	irqreturn_t ret = IRQ_NONE;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(dmc620_pmu, &irq->pmus_node, pmus_node) {
+		unsigned long status;
+		struct perf_event *event;
+		unsigned int idx;
+
+		/*
+		 * HW doesn't provide a control to atomically disable all counters.
+		 * To prevent race condition (overflow happens while clearing status register),
+		 * disable all events before continuing
+		 */
+		for (idx = 0; idx < DMC620_PMU_MAX_COUNTERS; idx++) {
+			event = dmc620_pmu->events[idx];
+			if (!event)
+				continue;
+			dmc620_pmu_disable_counter(event);
+		}
+
+		status = readl(dmc620_pmu->base + DMC620_PMU_OVERFLOW_STATUS_CLKDIV2);
+		status |= (readl(dmc620_pmu->base + DMC620_PMU_OVERFLOW_STATUS_CLK) <<
+				DMC620_PMU_CLKDIV2_MAX_COUNTERS);
+		if (status) {
+			for_each_set_bit(idx, &status,
+					DMC620_PMU_MAX_COUNTERS) {
+				event = dmc620_pmu->events[idx];
+				if (WARN_ON_ONCE(!event))
+					continue;
+				dmc620_pmu_event_update(event);
+				dmc620_pmu_event_set_period(event);
+			}
+
+			if (status & DMC620_PMU_OVERFLOW_STATUS_CLKDIV2_MASK)
+				writel(0, dmc620_pmu->base + DMC620_PMU_OVERFLOW_STATUS_CLKDIV2);
+
+			if ((status >> DMC620_PMU_CLKDIV2_MAX_COUNTERS) &
+				DMC620_PMU_OVERFLOW_STATUS_CLK_MASK)
+				writel(0, dmc620_pmu->base + DMC620_PMU_OVERFLOW_STATUS_CLK);
+		}
+
+		for (idx = 0; idx < DMC620_PMU_MAX_COUNTERS; idx++) {
+			event = dmc620_pmu->events[idx];
+			if (!event)
+				continue;
+			if (!(event->hw.state & PERF_HES_STOPPED))
+				dmc620_pmu_enable_counter(event);
+		}
+
+		ret = IRQ_HANDLED;
+	}
+	rcu_read_unlock();
+
+	return ret;
+}
+
+static struct dmc620_pmu_irq *__dmc620_pmu_get_irq(int irq_num)
+{
+	struct dmc620_pmu_irq *irq;
+	int ret;
+
+	list_for_each_entry(irq, &dmc620_pmu_irqs, irqs_node)
+		if (irq->irq_num == irq_num && refcount_inc_not_zero(&irq->refcount))
+			return irq;
+
+	irq = kzalloc(sizeof(*irq), GFP_KERNEL);
+	if (!irq)
+		return ERR_PTR(-ENOMEM);
+
+	INIT_LIST_HEAD(&irq->pmus_node);
+
+	/* Pick one CPU to be the preferred one to use */
+	irq->cpu = raw_smp_processor_id();
+	refcount_set(&irq->refcount, 1);
+
+	ret = request_irq(irq_num, dmc620_pmu_handle_irq,
+			  IRQF_NOBALANCING | IRQF_NO_THREAD,
+			  "dmc620-pmu", irq);
+	if (ret)
+		goto out_free_aff;
+
+	ret = irq_set_affinity_hint(irq_num, cpumask_of(irq->cpu));
+	if (ret)
+		goto out_free_irq;
+
+	ret = cpuhp_state_add_instance_nocalls(cpuhp_state_num, &irq->node);
+	if (ret)
+		goto out_free_irq;
+
+	irq->irq_num = irq_num;
+	list_add(&irq->irqs_node, &dmc620_pmu_irqs);
+
+	return irq;
+
+out_free_irq:
+	free_irq(irq_num, irq);
+out_free_aff:
+	kfree(irq);
+	return ERR_PTR(ret);
+}
+
+static int dmc620_pmu_get_irq(struct dmc620_pmu *dmc620_pmu, int irq_num)
+{
+	struct dmc620_pmu_irq *irq;
+
+	mutex_lock(&dmc620_pmu_irqs_lock);
+	irq = __dmc620_pmu_get_irq(irq_num);
+	mutex_unlock(&dmc620_pmu_irqs_lock);
+
+	if (IS_ERR(irq))
+		return PTR_ERR(irq);
+
+	dmc620_pmu->irq = irq;
+	mutex_lock(&dmc620_pmu_irqs_lock);
+	list_add_rcu(&dmc620_pmu->pmus_node, &irq->pmus_node);
+	mutex_unlock(&dmc620_pmu_irqs_lock);
+
+	return 0;
+}
+
+static void dmc620_pmu_put_irq(struct dmc620_pmu *dmc620_pmu)
+{
+	struct dmc620_pmu_irq *irq = dmc620_pmu->irq;
+
+	mutex_lock(&dmc620_pmu_irqs_lock);
+	list_del_rcu(&dmc620_pmu->pmus_node);
+
+	if (!refcount_dec_and_test(&irq->refcount)) {
+		mutex_unlock(&dmc620_pmu_irqs_lock);
+		return;
+	}
+
+	list_del(&irq->irqs_node);
+	mutex_unlock(&dmc620_pmu_irqs_lock);
+
+	WARN_ON(irq_set_affinity_hint(irq->irq_num, NULL));
+	free_irq(irq->irq_num, irq);
+	cpuhp_state_remove_instance_nocalls(cpuhp_state_num, &irq->node);
+	kfree(irq);
+}
+
+static int dmc620_pmu_event_init(struct perf_event *event)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+	struct hw_perf_event *hwc = &event->hw;
+	struct perf_event *sibling;
+
+	if (event->attr.type != event->pmu->type)
+		return -ENOENT;
+
+	/*
+	 * DMC 620 PMUs are shared across all cpus and cannot
+	 * support task bound and sampling events.
+	 */
+	if (is_sampling_event(event) ||
+		event->attach_state & PERF_ATTACH_TASK) {
+		dev_dbg(dmc620_pmu->pmu.dev,
+			"Can't support per-task counters\n");
+		return -EOPNOTSUPP;
+	}
+
+	/*
+	 * Many perf core operations (eg. events rotation) operate on a
+	 * single CPU context. This is obvious for CPU PMUs, where one
+	 * expects the same sets of events being observed on all CPUs,
+	 * but can lead to issues for off-core PMUs, where each
+	 * event could be theoretically assigned to a different CPU. To
+	 * mitigate this, we enforce CPU assignment to one, selected
+	 * processor.
+	 */
+	event->cpu = dmc620_pmu->irq->cpu;
+	if (event->cpu < 0)
+		return -EINVAL;
+
+	/*
+	 * We can't atomically disable all HW counters so only one event allowed,
+	 * although software events are acceptable.
+	 */
+	if (event->group_leader != event &&
+			!is_software_event(event->group_leader))
+		return -EINVAL;
+
+	for_each_sibling_event(sibling, event->group_leader) {
+		if (sibling != event &&
+				!is_software_event(sibling))
+			return -EINVAL;
+	}
+
+	hwc->idx = -1;
+	return 0;
+}
+
+static void dmc620_pmu_read(struct perf_event *event)
+{
+	dmc620_pmu_event_update(event);
+}
+
+static void dmc620_pmu_start(struct perf_event *event, int flags)
+{
+	event->hw.state = 0;
+	dmc620_pmu_event_set_period(event);
+	dmc620_pmu_enable_counter(event);
+}
+
+static void dmc620_pmu_stop(struct perf_event *event, int flags)
+{
+	if (event->hw.state & PERF_HES_STOPPED)
+		return;
+
+	dmc620_pmu_disable_counter(event);
+	dmc620_pmu_event_update(event);
+	event->hw.state |= PERF_HES_STOPPED | PERF_HES_UPTODATE;
+}
+
+static int dmc620_pmu_add(struct perf_event *event, int flags)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+	struct perf_event_attr *attr = &event->attr;
+	struct hw_perf_event *hwc = &event->hw;
+	int idx;
+	u64 reg;
+
+	idx = dmc620_get_event_idx(event);
+	if (idx < 0)
+		return idx;
+
+	hwc->idx = idx;
+	dmc620_pmu->events[idx] = event;
+	hwc->state = PERF_HES_STOPPED | PERF_HES_UPTODATE;
+
+	reg = ATTR_CFG_GET_FLD(attr, mask);
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      idx, DMC620_PMU_COUNTERn_MASK_31_00, lower_32_bits(reg));
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      idx, DMC620_PMU_COUNTERn_MASK_63_32, upper_32_bits(reg));
+
+	reg = ATTR_CFG_GET_FLD(attr, match);
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      idx, DMC620_PMU_COUNTERn_MATCH_31_00, lower_32_bits(reg));
+	dmc620_pmu_creg_write(dmc620_pmu,
+			      idx, DMC620_PMU_COUNTERn_MATCH_63_32, upper_32_bits(reg));
+
+	if (flags & PERF_EF_START)
+		dmc620_pmu_start(event, PERF_EF_RELOAD);
+
+	perf_event_update_userpage(event);
+	return 0;
+}
+
+static void dmc620_pmu_del(struct perf_event *event, int flags)
+{
+	struct dmc620_pmu *dmc620_pmu = to_dmc620_pmu(event->pmu);
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	dmc620_pmu_stop(event, PERF_EF_UPDATE);
+	dmc620_pmu->events[idx] = NULL;
+	clear_bit(idx, dmc620_pmu->used_mask);
+	perf_event_update_userpage(event);
+}
+
+static int dmc620_pmu_cpu_teardown(unsigned int cpu,
+				   struct hlist_node *node)
+{
+	struct dmc620_pmu_irq *irq;
+	struct dmc620_pmu *dmc620_pmu;
+	unsigned int target;
+
+	irq = hlist_entry_safe(node, struct dmc620_pmu_irq, node);
+	if (cpu != irq->cpu)
+		return 0;
+
+	target = cpumask_any_but(cpu_online_mask, cpu);
+	if (target >= nr_cpu_ids)
+		return 0;
+
+	/* We're only reading, but this isn't the place to be involving RCU */
+	mutex_lock(&dmc620_pmu_irqs_lock);
+	list_for_each_entry(dmc620_pmu, &irq->pmus_node, pmus_node)
+		perf_pmu_migrate_context(&dmc620_pmu->pmu, irq->cpu, target);
+	mutex_unlock(&dmc620_pmu_irqs_lock);
+
+	WARN_ON(irq_set_affinity_hint(irq->irq_num, cpumask_of(target)));
+	irq->cpu = target;
+
+	return 0;
+}
+
+static int dmc620_pmu_device_probe(struct platform_device *pdev)
+{
+	struct dmc620_pmu *dmc620_pmu;
+	struct resource *res;
+	char *name;
+	int irq_num;
+	int i, ret;
+
+	dmc620_pmu = devm_kzalloc(&pdev->dev,
+			sizeof(struct dmc620_pmu), GFP_KERNEL);
+	if (!dmc620_pmu)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, dmc620_pmu);
+
+	dmc620_pmu->pmu = (struct pmu) {
+		.module = THIS_MODULE,
+		.capabilities	= PERF_PMU_CAP_NO_EXCLUDE,
+		.task_ctx_nr	= perf_invalid_context,
+		.event_init	= dmc620_pmu_event_init,
+		.add		= dmc620_pmu_add,
+		.del		= dmc620_pmu_del,
+		.start		= dmc620_pmu_start,
+		.stop		= dmc620_pmu_stop,
+		.read		= dmc620_pmu_read,
+		.attr_groups	= dmc620_pmu_attr_groups,
+	};
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dmc620_pmu->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(dmc620_pmu->base))
+		return PTR_ERR(dmc620_pmu->base);
+
+	/* Make sure device is reset before enabling interrupt */
+	for (i = 0; i < DMC620_PMU_MAX_COUNTERS; i++)
+		dmc620_pmu_creg_write(dmc620_pmu, i, DMC620_PMU_COUNTERn_CONTROL, 0);
+	writel(0, dmc620_pmu->base + DMC620_PMU_OVERFLOW_STATUS_CLKDIV2);
+	writel(0, dmc620_pmu->base + DMC620_PMU_OVERFLOW_STATUS_CLK);
+
+	irq_num = platform_get_irq(pdev, 0);
+	if (irq_num < 0)
+		return irq_num;
+
+	ret = dmc620_pmu_get_irq(dmc620_pmu, irq_num);
+	if (ret)
+		return ret;
+
+	name = devm_kasprintf(&pdev->dev, GFP_KERNEL,
+				  "%s_%llx", DMC620_PMUNAME,
+				  (u64)(res->start >> DMC620_PA_SHIFT));
+	if (!name) {
+		dev_err(&pdev->dev,
+			  "Create name failed, PMU @%pa\n", &res->start);
+		goto out_teardown_dev;
+	}
+
+	ret = perf_pmu_register(&dmc620_pmu->pmu, name, -1);
+	if (ret)
+		goto out_teardown_dev;
+
+	return 0;
+
+out_teardown_dev:
+	dmc620_pmu_put_irq(dmc620_pmu);
+	synchronize_rcu();
+	return ret;
+}
+
+static int dmc620_pmu_device_remove(struct platform_device *pdev)
+{
+	struct dmc620_pmu *dmc620_pmu = platform_get_drvdata(pdev);
+
+	dmc620_pmu_put_irq(dmc620_pmu);
+
+	/* perf will synchronise RCU before devres can free dmc620_pmu */
+	perf_pmu_unregister(&dmc620_pmu->pmu);
+
+	return 0;
+}
+
+static const struct acpi_device_id dmc620_acpi_match[] = {
+	{ "ARMHD620", 0},
+	{},
+};
+MODULE_DEVICE_TABLE(acpi, dmc620_acpi_match);
+static struct platform_driver dmc620_pmu_driver = {
+	.driver	= {
+		.name		= DMC620_DRVNAME,
+		.acpi_match_table = dmc620_acpi_match,
+	},
+	.probe	= dmc620_pmu_device_probe,
+	.remove	= dmc620_pmu_device_remove,
+};
+
+static int __init dmc620_pmu_init(void)
+{
+	cpuhp_state_num = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN,
+				      DMC620_DRVNAME,
+				      NULL,
+				      dmc620_pmu_cpu_teardown);
+	if (cpuhp_state_num < 0)
+		return cpuhp_state_num;
+
+	return platform_driver_register(&dmc620_pmu_driver);
+}
+
+static void __exit dmc620_pmu_exit(void)
+{
+	platform_driver_unregister(&dmc620_pmu_driver);
+	cpuhp_remove_multi_state(cpuhp_state_num);
+}
+
+module_init(dmc620_pmu_init);
+module_exit(dmc620_pmu_exit);
+
+MODULE_DESCRIPTION("Perf driver for the ARM DMC-620 memory controller");
+MODULE_AUTHOR("Tuan Phan <tuanphan@os.amperecomputing.com");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/perf/arm_dsu_pmu.c b/drivers/perf/arm_dsu_pmu.c
index 98e68ed7db85..0459a3403469 100644
--- a/drivers/perf/arm_dsu_pmu.c
+++ b/drivers/perf/arm_dsu_pmu.c
@@ -716,9 +716,6 @@ static int dsu_pmu_device_probe(struct platform_device *pdev)
 	if (IS_ERR(dsu_pmu))
 		return PTR_ERR(dsu_pmu);
 
-	if (IS_ERR_OR_NULL(fwnode))
-		return -ENOENT;
-
 	if (is_of_node(fwnode))
 		rc = dsu_pmu_dt_get_cpus(&pdev->dev, &dsu_pmu->associated_cpus);
 	else if (is_acpi_device_node(fwnode))
diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
index cb2f55f450e4..794a37d50853 100644
--- a/drivers/perf/arm_pmu.c
+++ b/drivers/perf/arm_pmu.c
@@ -726,6 +726,11 @@ static int armpmu_get_cpu_irq(struct arm_pmu *pmu, int cpu)
 	return per_cpu(hw_events->irq, cpu);
 }
 
+bool arm_pmu_irq_is_nmi(void)
+{
+	return has_nmi;
+}
+
 /*
  * PMU hardware loses all context when a CPU goes offline.
  * When a CPU is hotplugged back in, since some hardware registers are
diff --git a/drivers/perf/arm_smmuv3_pmu.c b/drivers/perf/arm_smmuv3_pmu.c
index 5274f7fe359e..74474bb322c3 100644
--- a/drivers/perf/arm_smmuv3_pmu.c
+++ b/drivers/perf/arm_smmuv3_pmu.c
@@ -74,6 +74,7 @@
 #define SMMU_PMCG_CFGR_NCTR             GENMASK(5, 0)
 #define SMMU_PMCG_CR                    0xE04
 #define SMMU_PMCG_CR_ENABLE             BIT(0)
+#define SMMU_PMCG_IIDR                  0xE08
 #define SMMU_PMCG_CEID0                 0xE20
 #define SMMU_PMCG_CEID1                 0xE28
 #define SMMU_PMCG_IRQ_CTRL              0xE50
@@ -112,6 +113,7 @@ struct smmu_pmu {
 	void __iomem *reloc_base;
 	u64 counter_mask;
 	u32 options;
+	u32 iidr;
 	bool global_filter;
 };
 
@@ -552,6 +554,40 @@ static struct attribute_group smmu_pmu_events_group = {
 	.is_visible = smmu_pmu_event_is_visible,
 };
 
+static ssize_t smmu_pmu_identifier_attr_show(struct device *dev,
+					struct device_attribute *attr,
+					char *page)
+{
+	struct smmu_pmu *smmu_pmu = to_smmu_pmu(dev_get_drvdata(dev));
+
+	return snprintf(page, PAGE_SIZE, "0x%08x\n", smmu_pmu->iidr);
+}
+
+static umode_t smmu_pmu_identifier_attr_visible(struct kobject *kobj,
+						struct attribute *attr,
+						int n)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct smmu_pmu *smmu_pmu = to_smmu_pmu(dev_get_drvdata(dev));
+
+	if (!smmu_pmu->iidr)
+		return 0;
+	return attr->mode;
+}
+
+static struct device_attribute smmu_pmu_identifier_attr =
+	__ATTR(identifier, 0444, smmu_pmu_identifier_attr_show, NULL);
+
+static struct attribute *smmu_pmu_identifier_attrs[] = {
+	&smmu_pmu_identifier_attr.attr,
+	NULL
+};
+
+static struct attribute_group smmu_pmu_identifier_group = {
+	.attrs = smmu_pmu_identifier_attrs,
+	.is_visible = smmu_pmu_identifier_attr_visible,
+};
+
 /* Formats */
 PMU_FORMAT_ATTR(event,		   "config:0-15");
 PMU_FORMAT_ATTR(filter_stream_id,  "config1:0-31");
@@ -575,6 +611,7 @@ static const struct attribute_group *smmu_pmu_attr_grps[] = {
 	&smmu_pmu_cpumask_group,
 	&smmu_pmu_events_group,
 	&smmu_pmu_format_group,
+	&smmu_pmu_identifier_group,
 	NULL
 };
 
@@ -795,6 +832,8 @@ static int smmu_pmu_probe(struct platform_device *pdev)
 		return err;
 	}
 
+	smmu_pmu->iidr = readl_relaxed(smmu_pmu->reg_base + SMMU_PMCG_IIDR);
+
 	name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "smmuv3_pmcg_%llx",
 			      (res_0->start) >> SMMU_PMCG_PA_SHIFT);
 	if (!name) {
diff --git a/drivers/perf/fsl_imx8_ddr_perf.c b/drivers/perf/fsl_imx8_ddr_perf.c
index 397540a4b799..a11bfd8a0823 100644
--- a/drivers/perf/fsl_imx8_ddr_perf.c
+++ b/drivers/perf/fsl_imx8_ddr_perf.c
@@ -50,6 +50,7 @@ static DEFINE_IDA(ddr_ida);
 
 struct fsl_ddr_devtype_data {
 	unsigned int quirks;    /* quirks needed for different DDR Perf core */
+	const char *identifier;	/* system PMU identifier for userspace */
 };
 
 static const struct fsl_ddr_devtype_data imx8_devtype_data;
@@ -58,13 +59,32 @@ static const struct fsl_ddr_devtype_data imx8m_devtype_data = {
 	.quirks = DDR_CAP_AXI_ID_FILTER,
 };
 
+static const struct fsl_ddr_devtype_data imx8mq_devtype_data = {
+	.quirks = DDR_CAP_AXI_ID_FILTER,
+	.identifier = "i.MX8MQ",
+};
+
+static const struct fsl_ddr_devtype_data imx8mm_devtype_data = {
+	.quirks = DDR_CAP_AXI_ID_FILTER,
+	.identifier = "i.MX8MM",
+};
+
+static const struct fsl_ddr_devtype_data imx8mn_devtype_data = {
+	.quirks = DDR_CAP_AXI_ID_FILTER,
+	.identifier = "i.MX8MN",
+};
+
 static const struct fsl_ddr_devtype_data imx8mp_devtype_data = {
 	.quirks = DDR_CAP_AXI_ID_FILTER_ENHANCED,
+	.identifier = "i.MX8MP",
 };
 
 static const struct of_device_id imx_ddr_pmu_dt_ids[] = {
 	{ .compatible = "fsl,imx8-ddr-pmu", .data = &imx8_devtype_data},
 	{ .compatible = "fsl,imx8m-ddr-pmu", .data = &imx8m_devtype_data},
+	{ .compatible = "fsl,imx8mq-ddr-pmu", .data = &imx8mq_devtype_data},
+	{ .compatible = "fsl,imx8mm-ddr-pmu", .data = &imx8mm_devtype_data},
+	{ .compatible = "fsl,imx8mn-ddr-pmu", .data = &imx8mn_devtype_data},
 	{ .compatible = "fsl,imx8mp-ddr-pmu", .data = &imx8mp_devtype_data},
 	{ /* sentinel */ }
 };
@@ -84,6 +104,40 @@ struct ddr_pmu {
 	int id;
 };
 
+static ssize_t ddr_perf_identifier_show(struct device *dev,
+					struct device_attribute *attr,
+					char *page)
+{
+	struct ddr_pmu *pmu = dev_get_drvdata(dev);
+
+	return sprintf(page, "%s\n", pmu->devtype_data->identifier);
+}
+
+static umode_t ddr_perf_identifier_attr_visible(struct kobject *kobj,
+						struct attribute *attr,
+						int n)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct ddr_pmu *pmu = dev_get_drvdata(dev);
+
+	if (!pmu->devtype_data->identifier)
+		return 0;
+	return attr->mode;
+};
+
+static struct device_attribute ddr_perf_identifier_attr =
+	__ATTR(identifier, 0444, ddr_perf_identifier_show, NULL);
+
+static struct attribute *ddr_perf_identifier_attrs[] = {
+	&ddr_perf_identifier_attr.attr,
+	NULL,
+};
+
+static struct attribute_group ddr_perf_identifier_attr_group = {
+	.attrs = ddr_perf_identifier_attrs,
+	.is_visible = ddr_perf_identifier_attr_visible,
+};
+
 enum ddr_perf_filter_capabilities {
 	PERF_CAP_AXI_ID_FILTER = 0,
 	PERF_CAP_AXI_ID_FILTER_ENHANCED,
@@ -237,6 +291,7 @@ static const struct attribute_group *attr_groups[] = {
 	&ddr_perf_format_attr_group,
 	&ddr_perf_cpumask_attr_group,
 	&ddr_perf_filter_cap_attr_group,
+	&ddr_perf_identifier_attr_group,
 	NULL,
 };
 
@@ -361,25 +416,6 @@ static int ddr_perf_event_init(struct perf_event *event)
 	return 0;
 }
 
-
-static void ddr_perf_event_update(struct perf_event *event)
-{
-	struct ddr_pmu *pmu = to_ddr_pmu(event->pmu);
-	struct hw_perf_event *hwc = &event->hw;
-	u64 delta, prev_raw_count, new_raw_count;
-	int counter = hwc->idx;
-
-	do {
-		prev_raw_count = local64_read(&hwc->prev_count);
-		new_raw_count = ddr_perf_read_counter(pmu, counter);
-	} while (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
-			new_raw_count) != prev_raw_count);
-
-	delta = (new_raw_count - prev_raw_count) & 0xFFFFFFFF;
-
-	local64_add(delta, &event->count);
-}
-
 static void ddr_perf_counter_enable(struct ddr_pmu *pmu, int config,
 				  int counter, bool enable)
 {
@@ -404,6 +440,56 @@ static void ddr_perf_counter_enable(struct ddr_pmu *pmu, int config,
 	}
 }
 
+static bool ddr_perf_counter_overflow(struct ddr_pmu *pmu, int counter)
+{
+	int val;
+
+	val = readl_relaxed(pmu->base + counter * 4 + COUNTER_CNTL);
+
+	return val & CNTL_OVER;
+}
+
+static void ddr_perf_counter_clear(struct ddr_pmu *pmu, int counter)
+{
+	u8 reg = counter * 4 + COUNTER_CNTL;
+	int val;
+
+	val = readl_relaxed(pmu->base + reg);
+	val &= ~CNTL_CLEAR;
+	writel(val, pmu->base + reg);
+
+	val |= CNTL_CLEAR;
+	writel(val, pmu->base + reg);
+}
+
+static void ddr_perf_event_update(struct perf_event *event)
+{
+	struct ddr_pmu *pmu = to_ddr_pmu(event->pmu);
+	struct hw_perf_event *hwc = &event->hw;
+	u64 new_raw_count;
+	int counter = hwc->idx;
+	int ret;
+
+	new_raw_count = ddr_perf_read_counter(pmu, counter);
+	local64_add(new_raw_count, &event->count);
+
+	/*
+	 * For legacy SoCs: event counter continue counting when overflow,
+	 *                  no need to clear the counter.
+	 * For new SoCs: event counter stop counting when overflow, need
+	 *               clear counter to let it count again.
+	 */
+	if (counter != EVENT_CYCLES_COUNTER) {
+		ret = ddr_perf_counter_overflow(pmu, counter);
+		if (ret)
+			dev_warn_ratelimited(pmu->dev,  "events lost due to counter overflow (config 0x%llx)\n",
+					     event->attr.config);
+	}
+
+	/* clear counter every time for both cycle counter and event counter */
+	ddr_perf_counter_clear(pmu, counter);
+}
+
 static void ddr_perf_event_start(struct perf_event *event, int flags)
 {
 	struct ddr_pmu *pmu = to_ddr_pmu(event->pmu);
@@ -537,7 +623,7 @@ static irqreturn_t ddr_perf_irq_handler(int irq, void *p)
 {
 	int i;
 	struct ddr_pmu *pmu = (struct ddr_pmu *) p;
-	struct perf_event *event, *cycle_event = NULL;
+	struct perf_event *event;
 
 	/* all counter will stop if cycle counter disabled */
 	ddr_perf_counter_enable(pmu,
@@ -547,7 +633,9 @@ static irqreturn_t ddr_perf_irq_handler(int irq, void *p)
 	/*
 	 * When the cycle counter overflows, all counters are stopped,
 	 * and an IRQ is raised. If any other counter overflows, it
-	 * continues counting, and no IRQ is raised.
+	 * continues counting, and no IRQ is raised. But for new SoCs,
+	 * such as i.MX8MP, event counter would stop when overflow, so
+	 * we need use cycle counter to stop overflow of event counter.
 	 *
 	 * Cycles occur at least 4 times as often as other events, so we
 	 * can update all events on a cycle counter overflow and not
@@ -562,17 +650,12 @@ static irqreturn_t ddr_perf_irq_handler(int irq, void *p)
 		event = pmu->events[i];
 
 		ddr_perf_event_update(event);
-
-		if (event->hw.idx == EVENT_CYCLES_COUNTER)
-			cycle_event = event;
 	}
 
 	ddr_perf_counter_enable(pmu,
 			      EVENT_CYCLES_ID,
 			      EVENT_CYCLES_COUNTER,
 			      true);
-	if (cycle_event)
-		ddr_perf_event_update(cycle_event);
 
 	return IRQ_HANDLED;
 }
diff --git a/drivers/perf/hisilicon/hisi_uncore_ddrc_pmu.c b/drivers/perf/hisilicon/hisi_uncore_ddrc_pmu.c
index 5e3645c96443..5ac6c9113767 100644
--- a/drivers/perf/hisilicon/hisi_uncore_ddrc_pmu.c
+++ b/drivers/perf/hisilicon/hisi_uncore_ddrc_pmu.c
@@ -33,6 +33,7 @@
 #define DDRC_INT_MASK		0x6c8
 #define DDRC_INT_STATUS		0x6cc
 #define DDRC_INT_CLEAR		0x6d0
+#define DDRC_VERSION		0x710
 
 /* DDRC has 8-counters */
 #define DDRC_NR_COUNTERS	0x8
@@ -267,6 +268,8 @@ static int hisi_ddrc_pmu_init_data(struct platform_device *pdev,
 		return PTR_ERR(ddrc_pmu->base);
 	}
 
+	ddrc_pmu->identifier = readl(ddrc_pmu->base + DDRC_VERSION);
+
 	return 0;
 }
 
@@ -308,10 +311,23 @@ static const struct attribute_group hisi_ddrc_pmu_cpumask_attr_group = {
 	.attrs = hisi_ddrc_pmu_cpumask_attrs,
 };
 
+static struct device_attribute hisi_ddrc_pmu_identifier_attr =
+	__ATTR(identifier, 0444, hisi_uncore_pmu_identifier_attr_show, NULL);
+
+static struct attribute *hisi_ddrc_pmu_identifier_attrs[] = {
+	&hisi_ddrc_pmu_identifier_attr.attr,
+	NULL
+};
+
+static struct attribute_group hisi_ddrc_pmu_identifier_group = {
+	.attrs = hisi_ddrc_pmu_identifier_attrs,
+};
+
 static const struct attribute_group *hisi_ddrc_pmu_attr_groups[] = {
 	&hisi_ddrc_pmu_format_group,
 	&hisi_ddrc_pmu_events_group,
 	&hisi_ddrc_pmu_cpumask_attr_group,
+	&hisi_ddrc_pmu_identifier_group,
 	NULL,
 };
 
diff --git a/drivers/perf/hisilicon/hisi_uncore_hha_pmu.c b/drivers/perf/hisilicon/hisi_uncore_hha_pmu.c
index 5eb8168029c0..41b2dceb5f26 100644
--- a/drivers/perf/hisilicon/hisi_uncore_hha_pmu.c
+++ b/drivers/perf/hisilicon/hisi_uncore_hha_pmu.c
@@ -23,6 +23,7 @@
 #define HHA_INT_MASK		0x0804
 #define HHA_INT_STATUS		0x0808
 #define HHA_INT_CLEAR		0x080C
+#define HHA_VERSION		0x1cf0
 #define HHA_PERF_CTRL		0x1E00
 #define HHA_EVENT_CTRL		0x1E04
 #define HHA_EVENT_TYPE0		0x1E80
@@ -261,6 +262,8 @@ static int hisi_hha_pmu_init_data(struct platform_device *pdev,
 		return PTR_ERR(hha_pmu->base);
 	}
 
+	hha_pmu->identifier = readl(hha_pmu->base + HHA_VERSION);
+
 	return 0;
 }
 
@@ -320,10 +323,23 @@ static const struct attribute_group hisi_hha_pmu_cpumask_attr_group = {
 	.attrs = hisi_hha_pmu_cpumask_attrs,
 };
 
+static struct device_attribute hisi_hha_pmu_identifier_attr =
+	__ATTR(identifier, 0444, hisi_uncore_pmu_identifier_attr_show, NULL);
+
+static struct attribute *hisi_hha_pmu_identifier_attrs[] = {
+	&hisi_hha_pmu_identifier_attr.attr,
+	NULL
+};
+
+static struct attribute_group hisi_hha_pmu_identifier_group = {
+	.attrs = hisi_hha_pmu_identifier_attrs,
+};
+
 static const struct attribute_group *hisi_hha_pmu_attr_groups[] = {
 	&hisi_hha_pmu_format_group,
 	&hisi_hha_pmu_events_group,
 	&hisi_hha_pmu_cpumask_attr_group,
+	&hisi_hha_pmu_identifier_group,
 	NULL,
 };
 
diff --git a/drivers/perf/hisilicon/hisi_uncore_l3c_pmu.c b/drivers/perf/hisilicon/hisi_uncore_l3c_pmu.c
index 3e8b5eab5514..705501d18d03 100644
--- a/drivers/perf/hisilicon/hisi_uncore_l3c_pmu.c
+++ b/drivers/perf/hisilicon/hisi_uncore_l3c_pmu.c
@@ -25,6 +25,7 @@
 #define L3C_INT_STATUS		0x0808
 #define L3C_INT_CLEAR		0x080c
 #define L3C_EVENT_CTRL	        0x1c00
+#define L3C_VERSION		0x1cf0
 #define L3C_EVENT_TYPE0		0x1d00
 /*
  * Each counter is 48-bits and [48:63] are reserved
@@ -264,6 +265,8 @@ static int hisi_l3c_pmu_init_data(struct platform_device *pdev,
 		return PTR_ERR(l3c_pmu->base);
 	}
 
+	l3c_pmu->identifier = readl(l3c_pmu->base + L3C_VERSION);
+
 	return 0;
 }
 
@@ -310,10 +313,23 @@ static const struct attribute_group hisi_l3c_pmu_cpumask_attr_group = {
 	.attrs = hisi_l3c_pmu_cpumask_attrs,
 };
 
+static struct device_attribute hisi_l3c_pmu_identifier_attr =
+	__ATTR(identifier, 0444, hisi_uncore_pmu_identifier_attr_show, NULL);
+
+static struct attribute *hisi_l3c_pmu_identifier_attrs[] = {
+	&hisi_l3c_pmu_identifier_attr.attr,
+	NULL
+};
+
+static struct attribute_group hisi_l3c_pmu_identifier_group = {
+	.attrs = hisi_l3c_pmu_identifier_attrs,
+};
+
 static const struct attribute_group *hisi_l3c_pmu_attr_groups[] = {
 	&hisi_l3c_pmu_format_group,
 	&hisi_l3c_pmu_events_group,
 	&hisi_l3c_pmu_cpumask_attr_group,
+	&hisi_l3c_pmu_identifier_group,
 	NULL,
 };
 
diff --git a/drivers/perf/hisilicon/hisi_uncore_pmu.c b/drivers/perf/hisilicon/hisi_uncore_pmu.c
index 97aff877a4e7..9dbdc3fc3bb4 100644
--- a/drivers/perf/hisilicon/hisi_uncore_pmu.c
+++ b/drivers/perf/hisilicon/hisi_uncore_pmu.c
@@ -119,6 +119,16 @@ int hisi_uncore_pmu_get_event_idx(struct perf_event *event)
 }
 EXPORT_SYMBOL_GPL(hisi_uncore_pmu_get_event_idx);
 
+ssize_t hisi_uncore_pmu_identifier_attr_show(struct device *dev,
+					     struct device_attribute *attr,
+					     char *page)
+{
+	struct hisi_pmu *hisi_pmu = to_hisi_pmu(dev_get_drvdata(dev));
+
+	return snprintf(page, PAGE_SIZE, "0x%08x\n", hisi_pmu->identifier);
+}
+EXPORT_SYMBOL_GPL(hisi_uncore_pmu_identifier_attr_show);
+
 static void hisi_uncore_pmu_clear_event_idx(struct hisi_pmu *hisi_pmu, int idx)
 {
 	if (!hisi_uncore_pmu_counter_valid(hisi_pmu, idx)) {
diff --git a/drivers/perf/hisilicon/hisi_uncore_pmu.h b/drivers/perf/hisilicon/hisi_uncore_pmu.h
index b59ec22169ab..25b7cbe1f818 100644
--- a/drivers/perf/hisilicon/hisi_uncore_pmu.h
+++ b/drivers/perf/hisilicon/hisi_uncore_pmu.h
@@ -75,6 +75,7 @@ struct hisi_pmu {
 	int counter_bits;
 	/* check event code range */
 	int check_event;
+	u32 identifier;
 };
 
 int hisi_uncore_pmu_counter_valid(struct hisi_pmu *hisi_pmu, int idx);
@@ -97,4 +98,10 @@ ssize_t hisi_cpumask_sysfs_show(struct device *dev,
 				struct device_attribute *attr, char *buf);
 int hisi_uncore_pmu_online_cpu(unsigned int cpu, struct hlist_node *node);
 int hisi_uncore_pmu_offline_cpu(unsigned int cpu, struct hlist_node *node);
+
+ssize_t hisi_uncore_pmu_identifier_attr_show(struct device *dev,
+					     struct device_attribute *attr,
+					     char *page);
+
+
 #endif /* __HISI_UNCORE_PMU_H__ */