Merge tag 'perf_urgent_for_v6.16_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull perf fixes from Borislav Petkov:

 - Avoid a crash on a heterogeneous machine where not all cores support
   the same hw events features

 - Avoid a deadlock when throttling events

 - Document the perf event states more

 - Make sure a number of perf paths switching off or rescheduling events
   call perf_cgroup_event_disable()

 - Make sure perf does task sampling before its userspace mapping is
   torn down, and not after

* tag 'perf_urgent_for_v6.16_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/intel: Fix crash in icl_update_topdown_event()
  perf: Fix the throttle error of some clock events
  perf: Add comment to enum perf_event_state
  perf/core: Fix WARN in perf_cgroup_switch()
  perf: Fix dangling cgroup pointer in cpuctx
  perf: Fix cgroup state vs ERROR
  perf: Fix sample vs do_exit()

Author: torvalds

arch/x86/events/intel/core.c

@@ -2826,7 +2826,7 @@ static void intel_pmu_read_event(struct perf_event *event)
 		 * If the PEBS counters snapshotting is enabled,
 		 * the topdown event is available in PEBS records.
 		 */
-		if (is_topdown_event(event) && !is_pebs_counter_event_group(event))
+		if (is_topdown_count(event) && !is_pebs_counter_event_group(event))
 			static_call(intel_pmu_update_topdown_event)(event, NULL);
 		else
 			intel_pmu_drain_pebs_buffer();

include/linux/perf_event.h

@@ -635,8 +635,46 @@ struct perf_addr_filter_range {
 	unsigned long			size;
 };
 
-/**
- * enum perf_event_state - the states of an event:
+/*
+ * The normal states are:
+ *
+ *            ACTIVE    --.
+ *               ^        |
+ *               |        |
+ *       sched_{in,out}() |
+ *               |        |
+ *               v        |
+ *      ,---> INACTIVE  --+ <-.
+ *      |                 |   |
+ *      |                {dis,en}able()
+ *   sched_in()           |   |
+ *      |       OFF    <--' --+
+ *      |                     |
+ *      `--->  ERROR    ------'
+ *
+ * That is:
+ *
+ * sched_in:       INACTIVE          -> {ACTIVE,ERROR}
+ * sched_out:      ACTIVE            -> INACTIVE
+ * disable:        {ACTIVE,INACTIVE} -> OFF
+ * enable:         {OFF,ERROR}       -> INACTIVE
+ *
+ * Where {OFF,ERROR} are disabled states.
+ *
+ * Then we have the {EXIT,REVOKED,DEAD} states which are various shades of
+ * defunct events:
+ *
+ *  - EXIT means task that the even was assigned to died, but child events
+ *    still live, and further children can still be created. But the event
+ *    itself will never be active again. It can only transition to
+ *    {REVOKED,DEAD};
+ *
+ *  - REVOKED means the PMU the event was associated with is gone; all
+ *    functionality is stopped but the event is still alive. Can only
+ *    transition to DEAD;
+ *
+ *  - DEAD event really is DYING tearing down state and freeing bits.
+ *
  */
 enum perf_event_state {
 	PERF_EVENT_STATE_DEAD		= -5,

kernel/events/core.c

@@ -207,6 +207,19 @@ static void perf_ctx_unlock(struct perf_cpu_context *cpuctx,
 	__perf_ctx_unlock(&cpuctx->ctx);
 }
 
+typedef struct {
+	struct perf_cpu_context *cpuctx;
+	struct perf_event_context *ctx;
+} class_perf_ctx_lock_t;
+
+static inline void class_perf_ctx_lock_destructor(class_perf_ctx_lock_t *_T)
+{ perf_ctx_unlock(_T->cpuctx, _T->ctx); }
+
+static inline class_perf_ctx_lock_t
+class_perf_ctx_lock_constructor(struct perf_cpu_context *cpuctx,
+				struct perf_event_context *ctx)
+{ perf_ctx_lock(cpuctx, ctx); return (class_perf_ctx_lock_t){ cpuctx, ctx }; }
+
 #define TASK_TOMBSTONE ((void *)-1L)
 
 static bool is_kernel_event(struct perf_event *event)
@@ -944,7 +957,13 @@ static void perf_cgroup_switch(struct task_struct *task)
 	if (READ_ONCE(cpuctx->cgrp) == cgrp)
 		return;
 
-	perf_ctx_lock(cpuctx, cpuctx->task_ctx);
+	guard(perf_ctx_lock)(cpuctx, cpuctx->task_ctx);
+	/*
+	 * Re-check, could've raced vs perf_remove_from_context().
+	 */
+	if (READ_ONCE(cpuctx->cgrp) == NULL)
+		return;
+
 	perf_ctx_disable(&cpuctx->ctx, true);
 
 	ctx_sched_out(&cpuctx->ctx, NULL, EVENT_ALL|EVENT_CGROUP);
@@ -962,7 +981,6 @@ static void perf_cgroup_switch(struct task_struct *task)
 	ctx_sched_in(&cpuctx->ctx, NULL, EVENT_ALL|EVENT_CGROUP);
 
 	perf_ctx_enable(&cpuctx->ctx, true);
-	perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
 }
 
 static int perf_cgroup_ensure_storage(struct perf_event *event,
@@ -2120,18 +2138,6 @@ list_del_event(struct perf_event *event, struct perf_event_context *ctx)
 	if (event->group_leader == event)
 		del_event_from_groups(event, ctx);
 
-	/*
-	 * If event was in error state, then keep it
-	 * that way, otherwise bogus counts will be
-	 * returned on read(). The only way to get out
-	 * of error state is by explicit re-enabling
-	 * of the event
-	 */
-	if (event->state > PERF_EVENT_STATE_OFF) {
-		perf_cgroup_event_disable(event, ctx);
-		perf_event_set_state(event, PERF_EVENT_STATE_OFF);
-	}
-
 	ctx->generation++;
 	event->pmu_ctx->nr_events--;
 }
@@ -2149,8 +2155,9 @@ perf_aux_output_match(struct perf_event *event, struct perf_event *aux_event)
 }
 
 static void put_event(struct perf_event *event);
-static void event_sched_out(struct perf_event *event,
-			    struct perf_event_context *ctx);
+static void __event_disable(struct perf_event *event,
+			    struct perf_event_context *ctx,
+			    enum perf_event_state state);
 
 static void perf_put_aux_event(struct perf_event *event)
 {
@@ -2183,8 +2190,7 @@ static void perf_put_aux_event(struct perf_event *event)
 		 * state so that we don't try to schedule it again. Note
 		 * that perf_event_enable() will clear the ERROR status.
 		 */
-		event_sched_out(iter, ctx);
-		perf_event_set_state(event, PERF_EVENT_STATE_ERROR);
+		__event_disable(iter, ctx, PERF_EVENT_STATE_ERROR);
 	}
 }
 
@@ -2242,18 +2248,6 @@ static inline struct list_head *get_event_list(struct perf_event *event)
 				    &event->pmu_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)
-{
-	event_sched_out(event, event->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;
@@ -2289,8 +2283,15 @@ static void perf_group_detach(struct perf_event *event)
 	 */
 	list_for_each_entry_safe(sibling, tmp, &event->sibling_list, sibling_list) {
 
+		/*
+		 * 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.
+		 */
 		if (sibling->event_caps & PERF_EV_CAP_SIBLING)
-			perf_remove_sibling_event(sibling);
+			__event_disable(sibling, ctx, PERF_EVENT_STATE_ERROR);
 
 		sibling->group_leader = sibling;
 		list_del_init(&sibling->sibling_list);
@@ -2493,11 +2494,14 @@ __perf_remove_from_context(struct perf_event *event,
 		state = PERF_EVENT_STATE_EXIT;
 	if (flags & DETACH_REVOKE)
 		state = PERF_EVENT_STATE_REVOKED;
-	if (flags & DETACH_DEAD) {
-		event->pending_disable = 1;
+	if (flags & DETACH_DEAD)
 		state = PERF_EVENT_STATE_DEAD;
-	}
+
 	event_sched_out(event, ctx);
+
+	if (event->state > PERF_EVENT_STATE_OFF)
+		perf_cgroup_event_disable(event, ctx);
+
 	perf_event_set_state(event, min(event->state, state));
 
 	if (flags & DETACH_GROUP)
@@ -2562,6 +2566,15 @@ static void perf_remove_from_context(struct perf_event *event, unsigned long fla
 	event_function_call(event, __perf_remove_from_context, (void *)flags);
 }
 
+static void __event_disable(struct perf_event *event,
+			    struct perf_event_context *ctx,
+			    enum perf_event_state state)
+{
+	event_sched_out(event, ctx);
+	perf_cgroup_event_disable(event, ctx);
+	perf_event_set_state(event, state);
+}
+
 /*
  * Cross CPU call to disable a performance event
  */
@@ -2576,13 +2589,18 @@ static void __perf_event_disable(struct perf_event *event,
 	perf_pmu_disable(event->pmu_ctx->pmu);
 	ctx_time_update_event(ctx, event);
 
+	/*
+	 * When disabling a group leader, the whole group becomes ineligible
+	 * to run, so schedule out the full group.
+	 */
 	if (event == event->group_leader)
 		group_sched_out(event, ctx);
-	else
-		event_sched_out(event, ctx);
 
-	perf_event_set_state(event, PERF_EVENT_STATE_OFF);
-	perf_cgroup_event_disable(event, ctx);
+	/*
+	 * But only mark the leader OFF; the siblings will remain
+	 * INACTIVE.
+	 */
+	__event_disable(event, ctx, PERF_EVENT_STATE_OFF);
 
 	perf_pmu_enable(event->pmu_ctx->pmu);
 }
@@ -2656,8 +2674,8 @@ static void perf_event_unthrottle(struct perf_event *event, bool start)
 
 static void perf_event_throttle(struct perf_event *event)
 {
-	event->pmu->stop(event, 0);
 	event->hw.interrupts = MAX_INTERRUPTS;
+	event->pmu->stop(event, 0);
 	if (event == event->group_leader)
 		perf_log_throttle(event, 0);
 }
@@ -7439,6 +7457,10 @@ perf_sample_ustack_size(u16 stack_size, u16 header_size,
 	if (!regs)
 		return 0;
 
+	/* No mm, no stack, no dump. */
+	if (!current->mm)
+		return 0;
+
 	/*
 	 * Check if we fit in with the requested stack size into the:
 	 * - TASK_SIZE
@@ -8150,6 +8172,9 @@ perf_callchain(struct perf_event *event, struct pt_regs *regs)
 	const u32 max_stack = event->attr.sample_max_stack;
 	struct perf_callchain_entry *callchain;
 
+	if (!current->mm)
+		user = false;
+
 	if (!kernel && !user)
 		return &__empty_callchain;
 
@@ -11749,7 +11774,12 @@ static void perf_swevent_cancel_hrtimer(struct perf_event *event)
 {
 	struct hw_perf_event *hwc = &event->hw;
 
-	if (is_sampling_event(event)) {
+	/*
+	 * The throttle can be triggered in the hrtimer handler.
+	 * The HRTIMER_NORESTART should be used to stop the timer,
+	 * rather than hrtimer_cancel(). See perf_swevent_hrtimer()
+	 */
+	if (is_sampling_event(event) && (hwc->interrupts != MAX_INTERRUPTS)) {
 		ktime_t remaining = hrtimer_get_remaining(&hwc->hrtimer);
 		local64_set(&hwc->period_left, ktime_to_ns(remaining));
 
@@ -11804,7 +11834,8 @@ static void cpu_clock_event_start(struct perf_event *event, int flags)
 static void cpu_clock_event_stop(struct perf_event *event, int flags)
 {
 	perf_swevent_cancel_hrtimer(event);
-	cpu_clock_event_update(event);
+	if (flags & PERF_EF_UPDATE)
+		cpu_clock_event_update(event);
 }
 
 static int cpu_clock_event_add(struct perf_event *event, int flags)
@@ -11882,7 +11913,8 @@ static void task_clock_event_start(struct perf_event *event, int flags)
 static void task_clock_event_stop(struct perf_event *event, int flags)
 {
 	perf_swevent_cancel_hrtimer(event);
-	task_clock_event_update(event, event->ctx->time);
+	if (flags & PERF_EF_UPDATE)
+		task_clock_event_update(event, event->ctx->time);
 }
 
 static int task_clock_event_add(struct perf_event *event, int flags)

kernel/exit.c

@@ -940,6 +940,15 @@ void __noreturn do_exit(long code)
 	taskstats_exit(tsk, group_dead);
 	trace_sched_process_exit(tsk, group_dead);
 
+	/*
+	 * Since sampling can touch ->mm, make sure to stop everything before we
+	 * tear it down.
+	 *
+	 * Also flushes inherited counters to the parent - before the parent
+	 * gets woken up by child-exit notifications.
+	 */
+	perf_event_exit_task(tsk);
+
 	exit_mm();
 
 	if (group_dead)
@@ -955,14 +964,6 @@ void __noreturn do_exit(long code)
 	exit_task_work(tsk);
 	exit_thread(tsk);
 
-	/*
-	 * Flush inherited counters to the parent - before the parent
-	 * gets woken up by child-exit notifications.
-	 *
-	 * because of cgroup mode, must be called before cgroup_exit()
-	 */
-	perf_event_exit_task(tsk);
-
 	sched_autogroup_exit_task(tsk);
 	cgroup_exit(tsk);