Merge remote-tracking branch 'origin/master' into perf/core

Further perf/core patches will depend on:

  d3f7b1bb2040 ("mm/gup: fix gup_fast with dynamic page table folding")

which is already in Linus' tree.

10 files changed, 200 insertions, 98 deletions

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 7b802a778014..a88c94d65693 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -28,6 +28,7 @@
 #include <linux/bitops.h>
 #include <linux/device.h>
 #include <linux/nospec.h>
+#include <linux/static_call.h>
 
 #include <asm/apic.h>
 #include <asm/stacktrace.h>
@@ -52,6 +53,34 @@ DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events) = {
 DEFINE_STATIC_KEY_FALSE(rdpmc_never_available_key);
 DEFINE_STATIC_KEY_FALSE(rdpmc_always_available_key);
 
+/*
+ * This here uses DEFINE_STATIC_CALL_NULL() to get a static_call defined
+ * from just a typename, as opposed to an actual function.
+ */
+DEFINE_STATIC_CALL_NULL(x86_pmu_handle_irq,  *x86_pmu.handle_irq);
+DEFINE_STATIC_CALL_NULL(x86_pmu_disable_all, *x86_pmu.disable_all);
+DEFINE_STATIC_CALL_NULL(x86_pmu_enable_all,  *x86_pmu.enable_all);
+DEFINE_STATIC_CALL_NULL(x86_pmu_enable,	     *x86_pmu.enable);
+DEFINE_STATIC_CALL_NULL(x86_pmu_disable,     *x86_pmu.disable);
+
+DEFINE_STATIC_CALL_NULL(x86_pmu_add,  *x86_pmu.add);
+DEFINE_STATIC_CALL_NULL(x86_pmu_del,  *x86_pmu.del);
+DEFINE_STATIC_CALL_NULL(x86_pmu_read, *x86_pmu.read);
+
+DEFINE_STATIC_CALL_NULL(x86_pmu_schedule_events,       *x86_pmu.schedule_events);
+DEFINE_STATIC_CALL_NULL(x86_pmu_get_event_constraints, *x86_pmu.get_event_constraints);
+DEFINE_STATIC_CALL_NULL(x86_pmu_put_event_constraints, *x86_pmu.put_event_constraints);
+
+DEFINE_STATIC_CALL_NULL(x86_pmu_start_scheduling,  *x86_pmu.start_scheduling);
+DEFINE_STATIC_CALL_NULL(x86_pmu_commit_scheduling, *x86_pmu.commit_scheduling);
+DEFINE_STATIC_CALL_NULL(x86_pmu_stop_scheduling,   *x86_pmu.stop_scheduling);
+
+DEFINE_STATIC_CALL_NULL(x86_pmu_sched_task,    *x86_pmu.sched_task);
+DEFINE_STATIC_CALL_NULL(x86_pmu_swap_task_ctx, *x86_pmu.swap_task_ctx);
+
+DEFINE_STATIC_CALL_NULL(x86_pmu_drain_pebs,   *x86_pmu.drain_pebs);
+DEFINE_STATIC_CALL_NULL(x86_pmu_pebs_aliases, *x86_pmu.pebs_aliases);
+
 u64 __read_mostly hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
 				[PERF_COUNT_HW_CACHE_OP_MAX]
@@ -663,7 +692,7 @@ static void x86_pmu_disable(struct pmu *pmu)
 	cpuc->enabled = 0;
 	barrier();
 
-	x86_pmu.disable_all();
+	static_call(x86_pmu_disable_all)();
 }
 
 void x86_pmu_enable_all(int added)
@@ -910,8 +939,7 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
 	if (cpuc->txn_flags & PERF_PMU_TXN_ADD)
 		n0 -= cpuc->n_txn;
 
-	if (x86_pmu.start_scheduling)
-		x86_pmu.start_scheduling(cpuc);
+	static_call_cond(x86_pmu_start_scheduling)(cpuc);
 
 	for (i = 0, wmin = X86_PMC_IDX_MAX, wmax = 0; i < n; i++) {
 		c = cpuc->event_constraint[i];
@@ -928,7 +956,7 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
 		 * change due to external factors (sibling state, allow_tfa).
 		 */
 		if (!c || (c->flags & PERF_X86_EVENT_DYNAMIC)) {
-			c = x86_pmu.get_event_constraints(cpuc, i, cpuc->event_list[i]);
+			c = static_call(x86_pmu_get_event_constraints)(cpuc, i, cpuc->event_list[i]);
 			cpuc->event_constraint[i] = c;
 		}
 
@@ -1011,8 +1039,7 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
 	if (!unsched && assign) {
 		for (i = 0; i < n; i++) {
 			e = cpuc->event_list[i];
-			if (x86_pmu.commit_scheduling)
-				x86_pmu.commit_scheduling(cpuc, i, assign[i]);
+			static_call_cond(x86_pmu_commit_scheduling)(cpuc, i, assign[i]);
 		}
 	} else {
 		for (i = n0; i < n; i++) {
@@ -1021,15 +1048,13 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
 			/*
 			 * release events that failed scheduling
 			 */
-			if (x86_pmu.put_event_constraints)
-				x86_pmu.put_event_constraints(cpuc, e);
+			static_call_cond(x86_pmu_put_event_constraints)(cpuc, e);
 
 			cpuc->event_constraint[i] = NULL;
 		}
 	}
 
-	if (x86_pmu.stop_scheduling)
-		x86_pmu.stop_scheduling(cpuc);
+	static_call_cond(x86_pmu_stop_scheduling)(cpuc);
 
 	return unsched ? -EINVAL : 0;
 }
@@ -1267,7 +1292,7 @@ static void x86_pmu_enable(struct pmu *pmu)
 	cpuc->enabled = 1;
 	barrier();
 
-	x86_pmu.enable_all(added);
+	static_call(x86_pmu_enable_all)(added);
 }
 
 static DEFINE_PER_CPU(u64 [X86_PMC_IDX_MAX], pmc_prev_left);
@@ -1392,7 +1417,7 @@ static int x86_pmu_add(struct perf_event *event, int flags)
 	if (cpuc->txn_flags & PERF_PMU_TXN_ADD)
 		goto done_collect;
 
-	ret = x86_pmu.schedule_events(cpuc, n, assign);
+	ret = static_call(x86_pmu_schedule_events)(cpuc, n, assign);
 	if (ret)
 		goto out;
 	/*
@@ -1410,13 +1435,11 @@ done_collect:
 	cpuc->n_added += n - n0;
 	cpuc->n_txn += n - n0;
 
-	if (x86_pmu.add) {
-		/*
-		 * This is before x86_pmu_enable() will call x86_pmu_start(),
-		 * so we enable LBRs before an event needs them etc..
-		 */
-		x86_pmu.add(event);
-	}
+	/*
+	 * This is before x86_pmu_enable() will call x86_pmu_start(),
+	 * so we enable LBRs before an event needs them etc..
+	 */
+	static_call_cond(x86_pmu_add)(event);
 
 	ret = 0;
 out:
@@ -1444,7 +1467,7 @@ static void x86_pmu_start(struct perf_event *event, int flags)
 	cpuc->events[idx] = event;
 	__set_bit(idx, cpuc->active_mask);
 	__set_bit(idx, cpuc->running);
-	x86_pmu.enable(event);
+	static_call(x86_pmu_enable)(event);
 	perf_event_update_userpage(event);
 }
 
@@ -1514,7 +1537,7 @@ void x86_pmu_stop(struct perf_event *event, int flags)
 	struct hw_perf_event *hwc = &event->hw;
 
 	if (test_bit(hwc->idx, cpuc->active_mask)) {
-		x86_pmu.disable(event);
+		static_call(x86_pmu_disable)(event);
 		__clear_bit(hwc->idx, cpuc->active_mask);
 		cpuc->events[hwc->idx] = NULL;
 		WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED);
@@ -1564,8 +1587,7 @@ static void x86_pmu_del(struct perf_event *event, int flags)
 	if (i >= cpuc->n_events - cpuc->n_added)
 		--cpuc->n_added;
 
-	if (x86_pmu.put_event_constraints)
-		x86_pmu.put_event_constraints(cpuc, event);
+	static_call_cond(x86_pmu_put_event_constraints)(cpuc, event);
 
 	/* Delete the array entry. */
 	while (++i < cpuc->n_events) {
@@ -1580,13 +1602,12 @@ static void x86_pmu_del(struct perf_event *event, int flags)
 	perf_event_update_userpage(event);
 
 do_del:
-	if (x86_pmu.del) {
-		/*
-		 * This is after x86_pmu_stop(); so we disable LBRs after any
-		 * event can need them etc..
-		 */
-		x86_pmu.del(event);
-	}
+
+	/*
+	 * This is after x86_pmu_stop(); so we disable LBRs after any
+	 * event can need them etc..
+	 */
+	static_call_cond(x86_pmu_del)(event);
 }
 
 int x86_pmu_handle_irq(struct pt_regs *regs)
@@ -1664,7 +1685,7 @@ perf_event_nmi_handler(unsigned int cmd, struct pt_regs *regs)
 		return NMI_DONE;
 
 	start_clock = sched_clock();
-	ret = x86_pmu.handle_irq(regs);
+	ret = static_call(x86_pmu_handle_irq)(regs);
 	finish_clock = sched_clock();
 
 	perf_sample_event_took(finish_clock - start_clock);
@@ -1877,6 +1898,38 @@ ssize_t x86_event_sysfs_show(char *page, u64 config, u64 event)
 static struct attribute_group x86_pmu_attr_group;
 static struct attribute_group x86_pmu_caps_group;
 
+static void x86_pmu_static_call_update(void)
+{
+	static_call_update(x86_pmu_handle_irq, x86_pmu.handle_irq);
+	static_call_update(x86_pmu_disable_all, x86_pmu.disable_all);
+	static_call_update(x86_pmu_enable_all, x86_pmu.enable_all);
+	static_call_update(x86_pmu_enable, x86_pmu.enable);
+	static_call_update(x86_pmu_disable, x86_pmu.disable);
+
+	static_call_update(x86_pmu_add, x86_pmu.add);
+	static_call_update(x86_pmu_del, x86_pmu.del);
+	static_call_update(x86_pmu_read, x86_pmu.read);
+
+	static_call_update(x86_pmu_schedule_events, x86_pmu.schedule_events);
+	static_call_update(x86_pmu_get_event_constraints, x86_pmu.get_event_constraints);
+	static_call_update(x86_pmu_put_event_constraints, x86_pmu.put_event_constraints);
+
+	static_call_update(x86_pmu_start_scheduling, x86_pmu.start_scheduling);
+	static_call_update(x86_pmu_commit_scheduling, x86_pmu.commit_scheduling);
+	static_call_update(x86_pmu_stop_scheduling, x86_pmu.stop_scheduling);
+
+	static_call_update(x86_pmu_sched_task, x86_pmu.sched_task);
+	static_call_update(x86_pmu_swap_task_ctx, x86_pmu.swap_task_ctx);
+
+	static_call_update(x86_pmu_drain_pebs, x86_pmu.drain_pebs);
+	static_call_update(x86_pmu_pebs_aliases, x86_pmu.pebs_aliases);
+}
+
+static void _x86_pmu_read(struct perf_event *event)
+{
+	x86_perf_event_update(event);
+}
+
 static int __init init_hw_perf_events(void)
 {
 	struct x86_pmu_quirk *quirk;
@@ -1945,6 +1998,11 @@ static int __init init_hw_perf_events(void)
 	pr_info("... fixed-purpose events:   %d\n",     x86_pmu.num_counters_fixed);
 	pr_info("... event mask:             %016Lx\n", x86_pmu.intel_ctrl);
 
+	if (!x86_pmu.read)
+		x86_pmu.read = _x86_pmu_read;
+
+	x86_pmu_static_call_update();
+
 	/*
 	 * Install callbacks. Core will call them for each online
 	 * cpu.
@@ -1981,11 +2039,9 @@ out:
 }
 early_initcall(init_hw_perf_events);
 
-static inline void x86_pmu_read(struct perf_event *event)
+static void x86_pmu_read(struct perf_event *event)
 {
-	if (x86_pmu.read)
-		return x86_pmu.read(event);
-	x86_perf_event_update(event);
+	static_call(x86_pmu_read)(event);
 }
 
 /*
@@ -2066,7 +2122,7 @@ static int x86_pmu_commit_txn(struct pmu *pmu)
 	if (!x86_pmu_initialized())
 		return -EAGAIN;
 
-	ret = x86_pmu.schedule_events(cpuc, n, assign);
+	ret = static_call(x86_pmu_schedule_events)(cpuc, n, assign);
 	if (ret)
 		return ret;
 
@@ -2357,15 +2413,13 @@ static const struct attribute_group *x86_pmu_attr_groups[] = {
 
 static void x86_pmu_sched_task(struct perf_event_context *ctx, bool sched_in)
 {
-	if (x86_pmu.sched_task)
-		x86_pmu.sched_task(ctx, sched_in);
+	static_call_cond(x86_pmu_sched_task)(ctx, sched_in);
 }
 
 static void x86_pmu_swap_task_ctx(struct perf_event_context *prev,
 				  struct perf_event_context *next)
 {
-	if (x86_pmu.swap_task_ctx)
-		x86_pmu.swap_task_ctx(prev, next);
+	static_call_cond(x86_pmu_swap_task_ctx)(prev, next);
 }
 
 void perf_check_microcode(void)
diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index 0e590c57e312..546cc89217bb 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -2631,7 +2631,7 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status)
 		u64 pebs_enabled = cpuc->pebs_enabled;
 
 		handled++;
-		x86_pmu.drain_pebs(regs);
+		x86_pmu.drain_pebs(regs, &data);
 		status &= x86_pmu.intel_ctrl | GLOBAL_STATUS_TRACE_TOPAPMI;
 
 		/*
@@ -4988,6 +4988,12 @@ __init int intel_pmu_init(void)
 
 	x86_add_quirk(intel_arch_events_quirk); /* Install first, so it runs last */
 
+	if (version >= 5) {
+		x86_pmu.intel_cap.anythread_deprecated = edx.split.anythread_deprecated;
+		if (x86_pmu.intel_cap.anythread_deprecated)
+			pr_cont(" AnyThread deprecated, ");
+	}
+
 	/*
 	 * Install the hw-cache-events table:
 	 */
@@ -4999,7 +5005,7 @@ __init int intel_pmu_init(void)
 
 	case INTEL_FAM6_CORE2_MEROM:
 		x86_add_quirk(intel_clovertown_quirk);
-		/* fall through */
+		fallthrough;
 
 	case INTEL_FAM6_CORE2_MEROM_L:
 	case INTEL_FAM6_CORE2_PENRYN:
@@ -5380,7 +5386,7 @@ __init int intel_pmu_init(void)
 
 	case INTEL_FAM6_SKYLAKE_X:
 		pmem = true;
-		/* fall through */
+		fallthrough;
 	case INTEL_FAM6_SKYLAKE_L:
 	case INTEL_FAM6_SKYLAKE:
 	case INTEL_FAM6_KABYLAKE_L:
@@ -5432,7 +5438,7 @@ __init int intel_pmu_init(void)
 	case INTEL_FAM6_ICELAKE_X:
 	case INTEL_FAM6_ICELAKE_D:
 		pmem = true;
-		/* fall through */
+		fallthrough;
 	case INTEL_FAM6_ICELAKE_L:
 	case INTEL_FAM6_ICELAKE:
 	case INTEL_FAM6_TIGERLAKE_L:
@@ -5514,6 +5520,10 @@ __init int intel_pmu_init(void)
 	x86_pmu.intel_ctrl |=
 		((1LL << x86_pmu.num_counters_fixed)-1) << INTEL_PMC_IDX_FIXED;
 
+	/* AnyThread may be deprecated on arch perfmon v5 or later */
+	if (x86_pmu.intel_cap.anythread_deprecated)
+		x86_pmu.format_attrs = intel_arch_formats_attr;
+
 	if (x86_pmu.event_constraints) {
 		/*
 		 * event on fixed counter2 (REF_CYCLES) only works on this
diff --git a/arch/x86/events/intel/cstate.c b/arch/x86/events/intel/cstate.c
index a161a0bb4c42..407eee5f6f95 100644
--- a/arch/x86/events/intel/cstate.c
+++ b/arch/x86/events/intel/cstate.c
@@ -107,14 +107,14 @@
 MODULE_LICENSE("GPL");
 
 #define DEFINE_CSTATE_FORMAT_ATTR(_var, _name, _format)		\
-static ssize_t __cstate_##_var##_show(struct kobject *kobj,	\
-				struct kobj_attribute *attr,	\
+static ssize_t __cstate_##_var##_show(struct device *dev,	\
+				struct device_attribute *attr,	\
 				char *page)			\
 {								\
 	BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE);		\
 	return sprintf(page, _format "\n");			\
 }								\
-static struct kobj_attribute format_attr_##_var =		\
+static struct device_attribute format_attr_##_var =		\
 	__ATTR(_name, 0444, __cstate_##_var##_show, NULL)
 
 static ssize_t cstate_get_attr_cpumask(struct device *dev,
diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index 444e5f061d04..afbf734e2ee3 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -642,8 +642,8 @@ int intel_pmu_drain_bts_buffer(void)
 	rcu_read_lock();
 	perf_prepare_sample(&header, &data, event, &regs);
 
-	if (perf_output_begin(&handle, event, header.size *
-			      (top - base - skip)))
+	if (perf_output_begin(&handle, &data, event,
+			      header.size * (top - base - skip)))
 		goto unlock;
 
 	for (at = base; at < top; at++) {
@@ -670,7 +670,9 @@ unlock:
 
 static inline void intel_pmu_drain_pebs_buffer(void)
 {
-	x86_pmu.drain_pebs(NULL);
+	struct perf_sample_data data;
+
+	x86_pmu.drain_pebs(NULL, &data);
 }
 
 /*
@@ -1724,23 +1726,24 @@ intel_pmu_save_and_restart_reload(struct perf_event *event, int count)
 	return 0;
 }
 
-static void __intel_pmu_pebs_event(struct perf_event *event,
-				   struct pt_regs *iregs,
-				   void *base, void *top,
-				   int bit, int count,
-				   void (*setup_sample)(struct perf_event *,
-						struct pt_regs *,
-						void *,
-						struct perf_sample_data *,
-						struct pt_regs *))
+static __always_inline void
+__intel_pmu_pebs_event(struct perf_event *event,
+		       struct pt_regs *iregs,
+		       struct perf_sample_data *data,
+		       void *base, void *top,
+		       int bit, int count,
+		       void (*setup_sample)(struct perf_event *,
+					    struct pt_regs *,
+					    void *,
+					    struct perf_sample_data *,
+					    struct pt_regs *))
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct hw_perf_event *hwc = &event->hw;
-	struct perf_sample_data data;
 	struct x86_perf_regs perf_regs;
 	struct pt_regs *regs = &perf_regs.regs;
 	void *at = get_next_pebs_record_by_bit(base, top, bit);
-	struct pt_regs dummy_iregs;
+	static struct pt_regs dummy_iregs;
 
 	if (hwc->flags & PERF_X86_EVENT_AUTO_RELOAD) {
 		/*
@@ -1757,14 +1760,14 @@ static void __intel_pmu_pebs_event(struct perf_event *event,
 		iregs = &dummy_iregs;
 
 	while (count > 1) {
-		setup_sample(event, iregs, at, &data, regs);
-		perf_event_output(event, &data, regs);
+		setup_sample(event, iregs, at, data, regs);
+		perf_event_output(event, data, regs);
 		at += cpuc->pebs_record_size;
 		at = get_next_pebs_record_by_bit(at, top, bit);
 		count--;
 	}
 
-	setup_sample(event, iregs, at, &data, regs);
+	setup_sample(event, iregs, at, data, regs);
 	if (iregs == &dummy_iregs) {
 		/*
 		 * The PEBS records may be drained in the non-overflow context,
@@ -1772,18 +1775,18 @@ static void __intel_pmu_pebs_event(struct perf_event *event,
 		 * last record the same as other PEBS records, and doesn't
 		 * invoke the generic overflow handler.
 		 */
-		perf_event_output(event, &data, regs);
+		perf_event_output(event, data, regs);
 	} else {
 		/*
 		 * All but the last records are processed.
 		 * The last one is left to be able to call the overflow handler.
 		 */
-		if (perf_event_overflow(event, &data, regs))
+		if (perf_event_overflow(event, data, regs))
 			x86_pmu_stop(event, 0);
 	}
 }
 
-static void intel_pmu_drain_pebs_core(struct pt_regs *iregs)
+static void intel_pmu_drain_pebs_core(struct pt_regs *iregs, struct perf_sample_data *data)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct debug_store *ds = cpuc->ds;
@@ -1817,7 +1820,7 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs)
 		return;
 	}
 
-	__intel_pmu_pebs_event(event, iregs, at, top, 0, n,
+	__intel_pmu_pebs_event(event, iregs, data, at, top, 0, n,
 			       setup_pebs_fixed_sample_data);
 }
 
@@ -1840,7 +1843,7 @@ static void intel_pmu_pebs_event_update_no_drain(struct cpu_hw_events *cpuc, int
 	}
 }
 
-static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs)
+static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs, struct perf_sample_data *data)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct debug_store *ds = cpuc->ds;
@@ -1947,14 +1950,14 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs)
 		}
 
 		if (counts[bit]) {
-			__intel_pmu_pebs_event(event, iregs, base,
+			__intel_pmu_pebs_event(event, iregs, data, base,
 					       top, bit, counts[bit],
 					       setup_pebs_fixed_sample_data);
 		}
 	}
 }
 
-static void intel_pmu_drain_pebs_icl(struct pt_regs *iregs)
+static void intel_pmu_drain_pebs_icl(struct pt_regs *iregs, struct perf_sample_data *data)
 {
 	short counts[INTEL_PMC_IDX_FIXED + MAX_FIXED_PEBS_EVENTS] = {};
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
@@ -2002,7 +2005,7 @@ static void intel_pmu_drain_pebs_icl(struct pt_regs *iregs)
 		if (WARN_ON_ONCE(!event->attr.precise_ip))
 			continue;
 
-		__intel_pmu_pebs_event(event, iregs, base,
+		__intel_pmu_pebs_event(event, iregs, data, base,
 				       top, bit, counts[bit],
 				       setup_pebs_adaptive_sample_data);
 	}
diff --git a/arch/x86/events/intel/lbr.c b/arch/x86/events/intel/lbr.c
index 63f58bdf556c..8961653c5dd2 100644
--- a/arch/x86/events/intel/lbr.c
+++ b/arch/x86/events/intel/lbr.c
@@ -1268,7 +1268,7 @@ static int branch_type(unsigned long from, unsigned long to, int abort)
 			ret = X86_BR_ZERO_CALL;
 			break;
 		}
-		/* fall through */
+		fallthrough;
 	case 0x9a: /* call far absolute */
 		ret = X86_BR_CALL;
 		break;
diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c
index 1db6a7113bdb..357258f82dc8 100644
--- a/arch/x86/events/intel/uncore.c
+++ b/arch/x86/events/intel/uncore.c
@@ -94,8 +94,8 @@ end:
 	return map;
 }
 
-ssize_t uncore_event_show(struct kobject *kobj,
-			  struct kobj_attribute *attr, char *buf)
+ssize_t uncore_event_show(struct device *dev,
+			  struct device_attribute *attr, char *buf)
 {
 	struct uncore_event_desc *event =
 		container_of(attr, struct uncore_event_desc, attr);
diff --git a/arch/x86/events/intel/uncore.h b/arch/x86/events/intel/uncore.h
index 83d2a7d490e0..9efea154349d 100644
--- a/arch/x86/events/intel/uncore.h
+++ b/arch/x86/events/intel/uncore.h
@@ -157,7 +157,7 @@ struct intel_uncore_box {
 #define UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS	2
 
 struct uncore_event_desc {
-	struct kobj_attribute attr;
+	struct device_attribute attr;
 	const char *config;
 };
 
@@ -179,8 +179,8 @@ struct pci2phy_map {
 struct pci2phy_map *__find_pci2phy_map(int segment);
 int uncore_pcibus_to_physid(struct pci_bus *bus);
 
-ssize_t uncore_event_show(struct kobject *kobj,
-			  struct kobj_attribute *attr, char *buf);
+ssize_t uncore_event_show(struct device *dev,
+			  struct device_attribute *attr, char *buf);
 
 static inline struct intel_uncore_pmu *dev_to_uncore_pmu(struct device *dev)
 {
@@ -201,14 +201,14 @@ extern int __uncore_max_dies;
 }
 
 #define DEFINE_UNCORE_FORMAT_ATTR(_var, _name, _format)			\
-static ssize_t __uncore_##_var##_show(struct kobject *kobj,		\
-				struct kobj_attribute *attr,		\
+static ssize_t __uncore_##_var##_show(struct device *dev,		\
+				struct device_attribute *attr,		\
 				char *page)				\
 {									\
 	BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE);			\
 	return sprintf(page, _format "\n");				\
 }									\
-static struct kobj_attribute format_attr_##_var =			\
+static struct device_attribute format_attr_##_var =			\
 	__ATTR(_name, 0444, __uncore_##_var##_show, NULL)
 
 static inline bool uncore_pmc_fixed(int idx)
diff --git a/arch/x86/events/intel/uncore_snb.c b/arch/x86/events/intel/uncore_snb.c
index 6bbf54bf28b9..098f893e2e22 100644
--- a/arch/x86/events/intel/uncore_snb.c
+++ b/arch/x86/events/intel/uncore_snb.c
@@ -435,6 +435,18 @@ static struct uncore_event_desc snb_uncore_imc_events[] = {
 	INTEL_UNCORE_EVENT_DESC(data_writes.scale, "6.103515625e-5"),
 	INTEL_UNCORE_EVENT_DESC(data_writes.unit, "MiB"),
 
+	INTEL_UNCORE_EVENT_DESC(gt_requests, "event=0x03"),
+	INTEL_UNCORE_EVENT_DESC(gt_requests.scale, "6.103515625e-5"),
+	INTEL_UNCORE_EVENT_DESC(gt_requests.unit, "MiB"),
+
+	INTEL_UNCORE_EVENT_DESC(ia_requests, "event=0x04"),
+	INTEL_UNCORE_EVENT_DESC(ia_requests.scale, "6.103515625e-5"),
+	INTEL_UNCORE_EVENT_DESC(ia_requests.unit, "MiB"),
+
+	INTEL_UNCORE_EVENT_DESC(io_requests, "event=0x05"),
+	INTEL_UNCORE_EVENT_DESC(io_requests.scale, "6.103515625e-5"),
+	INTEL_UNCORE_EVENT_DESC(io_requests.unit, "MiB"),
+
 	{ /* end: all zeroes */ },
 };
 
@@ -450,13 +462,35 @@ static struct uncore_event_desc snb_uncore_imc_events[] = {
 #define SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE	0x5054
 #define SNB_UNCORE_PCI_IMC_CTR_BASE		SNB_UNCORE_PCI_IMC_DATA_READS_BASE
 
+/* BW break down- legacy counters */
+#define SNB_UNCORE_PCI_IMC_GT_REQUESTS		0x3
+#define SNB_UNCORE_PCI_IMC_GT_REQUESTS_BASE	0x5040
+#define SNB_UNCORE_PCI_IMC_IA_REQUESTS		0x4
+#define SNB_UNCORE_PCI_IMC_IA_REQUESTS_BASE	0x5044
+#define SNB_UNCORE_PCI_IMC_IO_REQUESTS		0x5
+#define SNB_UNCORE_PCI_IMC_IO_REQUESTS_BASE	0x5048
+
 enum perf_snb_uncore_imc_freerunning_types {
-	SNB_PCI_UNCORE_IMC_DATA		= 0,
+	SNB_PCI_UNCORE_IMC_DATA_READS		= 0,
+	SNB_PCI_UNCORE_IMC_DATA_WRITES,
+	SNB_PCI_UNCORE_IMC_GT_REQUESTS,
+	SNB_PCI_UNCORE_IMC_IA_REQUESTS,
+	SNB_PCI_UNCORE_IMC_IO_REQUESTS,
+
 	SNB_PCI_UNCORE_IMC_FREERUNNING_TYPE_MAX,
 };
 
 static struct freerunning_counters snb_uncore_imc_freerunning[] = {
-	[SNB_PCI_UNCORE_IMC_DATA]     = { SNB_UNCORE_PCI_IMC_DATA_READS_BASE, 0x4, 0x0, 2, 32 },
+	[SNB_PCI_UNCORE_IMC_DATA_READS]		= { SNB_UNCORE_PCI_IMC_DATA_READS_BASE,
+							0x0, 0x0, 1, 32 },
+	[SNB_PCI_UNCORE_IMC_DATA_WRITES]	= { SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE,
+							0x0, 0x0, 1, 32 },
+	[SNB_PCI_UNCORE_IMC_GT_REQUESTS]	= { SNB_UNCORE_PCI_IMC_GT_REQUESTS_BASE,
+							0x0, 0x0, 1, 32 },
+	[SNB_PCI_UNCORE_IMC_IA_REQUESTS]	= { SNB_UNCORE_PCI_IMC_IA_REQUESTS_BASE,
+							0x0, 0x0, 1, 32 },
+	[SNB_PCI_UNCORE_IMC_IO_REQUESTS]	= { SNB_UNCORE_PCI_IMC_IO_REQUESTS_BASE,
+							0x0, 0x0, 1, 32 },
 };
 
 static struct attribute *snb_uncore_imc_formats_attr[] = {
@@ -570,6 +604,18 @@ static int snb_uncore_imc_event_init(struct perf_event *event)
 		base = SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE;
 		idx = UNCORE_PMC_IDX_FREERUNNING;
 		break;
+	case SNB_UNCORE_PCI_IMC_GT_REQUESTS:
+		base = SNB_UNCORE_PCI_IMC_GT_REQUESTS_BASE;
+		idx = UNCORE_PMC_IDX_FREERUNNING;
+		break;
+	case SNB_UNCORE_PCI_IMC_IA_REQUESTS:
+		base = SNB_UNCORE_PCI_IMC_IA_REQUESTS_BASE;
+		idx = UNCORE_PMC_IDX_FREERUNNING;
+		break;
+	case SNB_UNCORE_PCI_IMC_IO_REQUESTS:
+		base = SNB_UNCORE_PCI_IMC_IO_REQUESTS_BASE;
+		idx = UNCORE_PMC_IDX_FREERUNNING;
+		break;
 	default:
 		return -EINVAL;
 	}
@@ -643,7 +689,7 @@ static struct intel_uncore_ops snb_uncore_imc_ops = {
 
 static struct intel_uncore_type snb_uncore_imc = {
 	.name		= "imc",
-	.num_counters   = 2,
+	.num_counters   = 5,
 	.num_boxes	= 1,
 	.num_freerunning_types	= SNB_PCI_UNCORE_IMC_FREERUNNING_TYPE_MAX,
 	.mmio_map_size	= SNB_UNCORE_PCI_IMC_MAP_SIZE,
diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
index 10032f023fcc..7895cf4c59a7 100644
--- a/arch/x86/events/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -585,6 +585,7 @@ union perf_capabilities {
 		u64     pebs_baseline:1;
 		u64	perf_metrics:1;
 		u64	pebs_output_pt_available:1;
+		u64	anythread_deprecated:1;
 	};
 	u64	capabilities;
 };
@@ -727,7 +728,7 @@ struct x86_pmu {
 	int		pebs_record_size;
 	int		pebs_buffer_size;
 	int		max_pebs_events;
-	void		(*drain_pebs)(struct pt_regs *regs);
+	void		(*drain_pebs)(struct pt_regs *regs, struct perf_sample_data *data);
 	struct event_constraint *pebs_constraints;
 	void		(*pebs_aliases)(struct perf_event *event);
 	unsigned long	large_pebs_flags;
diff --git a/arch/x86/events/rapl.c b/arch/x86/events/rapl.c
index 7c0120e2e957..7dbbeaacd995 100644
--- a/arch/x86/events/rapl.c
+++ b/arch/x86/events/rapl.c
@@ -93,18 +93,6 @@ static const char *const rapl_domain_names[NR_RAPL_DOMAINS] __initconst = {
  * any other bit is reserved
  */
 #define RAPL_EVENT_MASK	0xFFULL
-
-#define DEFINE_RAPL_FORMAT_ATTR(_var, _name, _format)		\
-static ssize_t __rapl_##_var##_show(struct kobject *kobj,	\
-				struct kobj_attribute *attr,	\
-				char *page)			\
-{								\
-	BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE);		\
-	return sprintf(page, _format "\n");			\
-}								\
-static struct kobj_attribute format_attr_##_var =		\
-	__ATTR(_name, 0444, __rapl_##_var##_show, NULL)
-
 #define RAPL_CNTR_WIDTH 32
 
 #define RAPL_EVENT_ATTR_STR(_name, v, str)					\
@@ -441,7 +429,7 @@ static struct attribute_group rapl_pmu_events_group = {
 	.attrs = attrs_empty,
 };
 
-DEFINE_RAPL_FORMAT_ATTR(event, event, "config:0-7");
+PMU_FORMAT_ATTR(event, "config:0-7");
 static struct attribute *rapl_formats_attr[] = {
 	&format_attr_event.attr,
 	NULL,