Merge tag 'perf-core-2020-10-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull performance events updates from Ingo Molnar:
 "x86 Intel updates:

   - Add Jasper Lake support

   - Add support for TopDown metrics on Ice Lake

   - Fix Ice Lake & Tiger Lake uncore support, add Snow Ridge support

   - Add a PCI sub driver to support uncore PMUs where the PCI resources
     have been claimed already - extending the range of supported
     systems.

  x86 AMD updates:

   - Restore 'perf stat -a' behaviour to program the uncore PMU to count
     all CPU threads.

   - Fix setting the proper count when sampling Large Increment per
     Cycle events / 'paired' events.

   - Fix IBS Fetch sampling on F17h and some other IBS fine tuning,
     greatly reducing the number of interrupts when large sample periods
     are specified.

   - Extends Family 17h RAPL support to also work on compatible F19h
     machines.

  Core code updates:

   - Fix race in perf_mmap_close()

   - Add PERF_EV_CAP_SIBLING, to denote that sibling events should be
     closed if the leader is removed.

   - Smaller fixes and updates"

* tag 'perf-core-2020-10-12' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (45 commits)
  perf/core: Fix race in the perf_mmap_close() function
  perf/x86: Fix n_metric for cancelled txn
  perf/x86: Fix n_pair for cancelled txn
  x86/events/amd/iommu: Fix sizeof mismatch
  perf/x86/intel: Check perf metrics feature for each CPU
  perf/x86/intel: Fix Ice Lake event constraint table
  perf/x86/intel/uncore: Fix the scale of the IMC free-running events
  perf/x86/intel/uncore: Fix for iio mapping on Skylake Server
  perf/x86/msr: Add Jasper Lake support
  perf/x86/intel: Add Jasper Lake support
  perf/x86/intel/uncore: Reduce the number of CBOX counters
  perf/x86/intel/uncore: Update Ice Lake uncore units
  perf/x86/intel/uncore: Split the Ice Lake and Tiger Lake MSR uncore support
  perf/x86/intel/uncore: Support PCIe3 unit on Snow Ridge
  perf/x86/intel/uncore: Generic support for the PCI sub driver
  perf/x86/intel/uncore: Factor out uncore_pci_pmu_unregister()
  perf/x86/intel/uncore: Factor out uncore_pci_pmu_register()
  perf/x86/intel/uncore: Factor out uncore_pci_find_dev_pmu()
  perf/x86/intel/uncore: Factor out uncore_pci_get_dev_die_info()
  perf/amd/uncore: Inform the user how many counters each uncore PMU has
  ...

kernel/events/core.c

@@ -383,7 +383,6 @@ static DEFINE_MUTEX(perf_sched_mutex);
 static atomic_t perf_sched_count;
 
 static DEFINE_PER_CPU(atomic_t, perf_cgroup_events);
-static DEFINE_PER_CPU(int, perf_sched_cb_usages);
 static DEFINE_PER_CPU(struct pmu_event_list, pmu_sb_events);
 
 static atomic_t nr_mmap_events __read_mostly;
@@ -2134,8 +2133,24 @@ static inline struct list_head *get_event_list(struct perf_event *event)
 	return event->attr.pinned ? &ctx->pinned_active : &ctx->flexible_active;
 }
 
+/*
+ * Events that have PERF_EV_CAP_SIBLING require being part of a group and
+ * cannot exist on their own, schedule them out and move them into the ERROR
+ * state. Also see _perf_event_enable(), it will not be able to recover
+ * this ERROR state.
+ */
+static inline void perf_remove_sibling_event(struct perf_event *event)
+{
+	struct perf_event_context *ctx = event->ctx;
+	struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);
+
+	event_sched_out(event, cpuctx, ctx);
+	perf_event_set_state(event, PERF_EVENT_STATE_ERROR);
+}
+
 static void perf_group_detach(struct perf_event *event)
 {
+	struct perf_event *leader = event->group_leader;
 	struct perf_event *sibling, *tmp;
 	struct perf_event_context *ctx = event->ctx;
 
@@ -2154,7 +2169,7 @@ static void perf_group_detach(struct perf_event *event)
 	/*
 	 * If this is a sibling, remove it from its group.
 	 */
-	if (event->group_leader != event) {
+	if (leader != event) {
 		list_del_init(&event->sibling_list);
 		event->group_leader->nr_siblings--;
 		goto out;
@@ -2167,6 +2182,9 @@ static void perf_group_detach(struct perf_event *event)
 	 */
 	list_for_each_entry_safe(sibling, tmp, &event->sibling_list, sibling_list) {
 
+		if (sibling->event_caps & PERF_EV_CAP_SIBLING)
+			perf_remove_sibling_event(sibling);
+
 		sibling->group_leader = sibling;
 		list_del_init(&sibling->sibling_list);
 
@@ -2184,10 +2202,10 @@ static void perf_group_detach(struct perf_event *event)
 	}
 
 out:
-	perf_event__header_size(event->group_leader);
-
-	for_each_sibling_event(tmp, event->group_leader)
+	for_each_sibling_event(tmp, leader)
 		perf_event__header_size(tmp);
+
+	perf_event__header_size(leader);
 }
 
 static bool is_orphaned_event(struct perf_event *event)
@@ -2980,6 +2998,7 @@ static void _perf_event_enable(struct perf_event *event)
 	raw_spin_lock_irq(&ctx->lock);
 	if (event->state >= PERF_EVENT_STATE_INACTIVE ||
 	    event->state <  PERF_EVENT_STATE_ERROR) {
+out:
 		raw_spin_unlock_irq(&ctx->lock);
 		return;
 	}
@@ -2991,8 +3010,16 @@ static void _perf_event_enable(struct perf_event *event)
 	 * has gone back into error state, as distinct from the task having
 	 * been scheduled away before the cross-call arrived.
 	 */
-	if (event->state == PERF_EVENT_STATE_ERROR)
+	if (event->state == PERF_EVENT_STATE_ERROR) {
+		/*
+		 * Detached SIBLING events cannot leave ERROR state.
+		 */
+		if (event->event_caps & PERF_EV_CAP_SIBLING &&
+		    event->group_leader == event)
+			goto out;
+
 		event->state = PERF_EVENT_STATE_OFF;
+	}
 	raw_spin_unlock_irq(&ctx->lock);
 
 	event_function_call(event, __perf_event_enable, NULL);
@@ -3357,10 +3384,12 @@ static void perf_event_context_sched_out(struct task_struct *task, int ctxn,
 	struct perf_event_context *parent, *next_parent;
 	struct perf_cpu_context *cpuctx;
 	int do_switch = 1;
+	struct pmu *pmu;
 
 	if (likely(!ctx))
 		return;
 
+	pmu = ctx->pmu;
 	cpuctx = __get_cpu_context(ctx);
 	if (!cpuctx->task_ctx)
 		return;
@@ -3390,11 +3419,15 @@ static void perf_event_context_sched_out(struct task_struct *task, int ctxn,
 		raw_spin_lock(&ctx->lock);
 		raw_spin_lock_nested(&next_ctx->lock, SINGLE_DEPTH_NESTING);
 		if (context_equiv(ctx, next_ctx)) {
-			struct pmu *pmu = ctx->pmu;
 
 			WRITE_ONCE(ctx->task, next);
 			WRITE_ONCE(next_ctx->task, task);
 
+			perf_pmu_disable(pmu);
+
+			if (cpuctx->sched_cb_usage && pmu->sched_task)
+				pmu->sched_task(ctx, false);
+
 			/*
 			 * PMU specific parts of task perf context can require
 			 * additional synchronization. As an example of such
@@ -3406,6 +3439,8 @@ static void perf_event_context_sched_out(struct task_struct *task, int ctxn,
 			else
 				swap(ctx->task_ctx_data, next_ctx->task_ctx_data);
 
+			perf_pmu_enable(pmu);
+
 			/*
 			 * RCU_INIT_POINTER here is safe because we've not
 			 * modified the ctx and the above modification of
@@ -3428,21 +3463,22 @@ unlock:
 
 	if (do_switch) {
 		raw_spin_lock(&ctx->lock);
+		perf_pmu_disable(pmu);
+
+		if (cpuctx->sched_cb_usage && pmu->sched_task)
+			pmu->sched_task(ctx, false);
 		task_ctx_sched_out(cpuctx, ctx, EVENT_ALL);
+
+		perf_pmu_enable(pmu);
 		raw_spin_unlock(&ctx->lock);
 	}
 }
 
-static DEFINE_PER_CPU(struct list_head, sched_cb_list);
-
 void perf_sched_cb_dec(struct pmu *pmu)
 {
 	struct perf_cpu_context *cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);
 
-	this_cpu_dec(perf_sched_cb_usages);
-
-	if (!--cpuctx->sched_cb_usage)
-		list_del(&cpuctx->sched_cb_entry);
+	--cpuctx->sched_cb_usage;
 }
 
 
@@ -3450,10 +3486,7 @@ void perf_sched_cb_inc(struct pmu *pmu)
 {
 	struct perf_cpu_context *cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);
 
-	if (!cpuctx->sched_cb_usage++)
-		list_add(&cpuctx->sched_cb_entry, this_cpu_ptr(&sched_cb_list));
-
-	this_cpu_inc(perf_sched_cb_usages);
+	cpuctx->sched_cb_usage++;
 }
 
 /*
@@ -3464,30 +3497,22 @@ void perf_sched_cb_inc(struct pmu *pmu)
  * PEBS requires this to provide PID/TID information. This requires we flush
  * all queued PEBS records before we context switch to a new task.
  */
-static void perf_pmu_sched_task(struct task_struct *prev,
-				struct task_struct *next,
-				bool sched_in)
+static void __perf_pmu_sched_task(struct perf_cpu_context *cpuctx, bool sched_in)
 {
-	struct perf_cpu_context *cpuctx;
 	struct pmu *pmu;
 
-	if (prev == next)
+	pmu = cpuctx->ctx.pmu; /* software PMUs will not have sched_task */
+
+	if (WARN_ON_ONCE(!pmu->sched_task))
 		return;
 
-	list_for_each_entry(cpuctx, this_cpu_ptr(&sched_cb_list), sched_cb_entry) {
-		pmu = cpuctx->ctx.pmu; /* software PMUs will not have sched_task */
+	perf_ctx_lock(cpuctx, cpuctx->task_ctx);
+	perf_pmu_disable(pmu);
 
-		if (WARN_ON_ONCE(!pmu->sched_task))
-			continue;
+	pmu->sched_task(cpuctx->task_ctx, sched_in);
 
-		perf_ctx_lock(cpuctx, cpuctx->task_ctx);
-		perf_pmu_disable(pmu);
-
-		pmu->sched_task(cpuctx->task_ctx, sched_in);
-
-		perf_pmu_enable(pmu);
-		perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
-	}
+	perf_pmu_enable(pmu);
+	perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
 }
 
 static void perf_event_switch(struct task_struct *task,
@@ -3512,9 +3537,6 @@ void __perf_event_task_sched_out(struct task_struct *task,
 {
 	int ctxn;
 
-	if (__this_cpu_read(perf_sched_cb_usages))
-		perf_pmu_sched_task(task, next, false);
-
 	if (atomic_read(&nr_switch_events))
 		perf_event_switch(task, next, false);
 
@@ -3746,10 +3768,14 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx,
 					struct task_struct *task)
 {
 	struct perf_cpu_context *cpuctx;
+	struct pmu *pmu = ctx->pmu;
 
 	cpuctx = __get_cpu_context(ctx);
-	if (cpuctx->task_ctx == ctx)
+	if (cpuctx->task_ctx == ctx) {
+		if (cpuctx->sched_cb_usage)
+			__perf_pmu_sched_task(cpuctx, true);
 		return;
+	}
 
 	perf_ctx_lock(cpuctx, ctx);
 	/*
@@ -3759,7 +3785,7 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx,
 	if (!ctx->nr_events)
 		goto unlock;
 
-	perf_pmu_disable(ctx->pmu);
+	perf_pmu_disable(pmu);
 	/*
 	 * We want to keep the following priority order:
 	 * cpu pinned (that don't need to move), task pinned,
@@ -3771,7 +3797,11 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx,
 	if (!RB_EMPTY_ROOT(&ctx->pinned_groups.tree))
 		cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE);
 	perf_event_sched_in(cpuctx, ctx, task);
-	perf_pmu_enable(ctx->pmu);
+
+	if (cpuctx->sched_cb_usage && pmu->sched_task)
+		pmu->sched_task(cpuctx->task_ctx, true);
+
+	perf_pmu_enable(pmu);
 
 unlock:
 	perf_ctx_unlock(cpuctx, ctx);
@@ -3814,9 +3844,6 @@ void __perf_event_task_sched_in(struct task_struct *prev,
 
 	if (atomic_read(&nr_switch_events))
 		perf_event_switch(task, prev, true);
-
-	if (__this_cpu_read(perf_sched_cb_usages))
-		perf_pmu_sched_task(prev, task, true);
 }
 
 static u64 perf_calculate_period(struct perf_event *event, u64 nsec, u64 count)
@@ -5869,11 +5896,11 @@ static void perf_pmu_output_stop(struct perf_event *event);
 static void perf_mmap_close(struct vm_area_struct *vma)
 {
 	struct perf_event *event = vma->vm_file->private_data;
-
 	struct perf_buffer *rb = ring_buffer_get(event);
 	struct user_struct *mmap_user = rb->mmap_user;
 	int mmap_locked = rb->mmap_locked;
 	unsigned long size = perf_data_size(rb);
+	bool detach_rest = false;
 
 	if (event->pmu->event_unmapped)
 		event->pmu->event_unmapped(event, vma->vm_mm);
@@ -5904,7 +5931,8 @@ static void perf_mmap_close(struct vm_area_struct *vma)
 		mutex_unlock(&event->mmap_mutex);
 	}
 
-	atomic_dec(&rb->mmap_count);
+	if (atomic_dec_and_test(&rb->mmap_count))
+		detach_rest = true;
 
 	if (!atomic_dec_and_mutex_lock(&event->mmap_count, &event->mmap_mutex))
 		goto out_put;
@@ -5913,7 +5941,7 @@ static void perf_mmap_close(struct vm_area_struct *vma)
 	mutex_unlock(&event->mmap_mutex);
 
 	/* If there's still other mmap()s of this buffer, we're done. */
-	if (atomic_read(&rb->mmap_count))
+	if (!detach_rest)
 		goto out_put;
 
 	/*
@@ -12829,7 +12857,6 @@ static void __init perf_event_init_all_cpus(void)
 #ifdef CONFIG_CGROUP_PERF
 		INIT_LIST_HEAD(&per_cpu(cgrp_cpuctx_list, cpu));
 #endif
-		INIT_LIST_HEAD(&per_cpu(sched_cb_list, cpu));
 	}
 }