[GIT PULL] perf fixes

[Ingo Molnar]

Linus,

Please pull the latest perf-urgent-for-linus git tree from:

   git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git perf-urgent-for-linus

   # HEAD: fff4b87e594ad3d2e4f51e8d3d86a6f9d3d8b654 perf/x86/intel/uncore: Make package handling more robust

Five kernel fixes:

 - an mmap tracing ABI fix for certain mappings
 - a use-after-free fix, found via KASAN
 - three CPU hotplug related x86 PMU driver fixes

 Thanks,

	Ingo

------------------>
Peter Zijlstra (2):
      perf/core: Fix use-after-free bug
      perf/core: Fix PERF_RECORD_MMAP2 prot/flags for anonymous memory

Thomas Gleixner (3):
      perf/x86/intel/rapl: Make package handling more robust
      perf/x86/intel/uncore: Clean up hotplug conversion fallout
      perf/x86/intel/uncore: Make package handling more robust


 arch/x86/events/intel/rapl.c   |  60 +++++------
 arch/x86/events/intel/uncore.c | 232 ++++++++++++++++-------------------------
 include/linux/cpuhotplug.h     |   3 -
 kernel/events/core.c           |  69 ++++++++----
 4 files changed, 163 insertions(+), 201 deletions(-)

diff --git a/arch/x86/events/intel/rapl.c b/arch/x86/events/intel/rapl.c
index 17c3564d087a..22ef4f72cf32 100644
--- a/arch/x86/events/intel/rapl.c
+++ b/arch/x86/events/intel/rapl.c
@@ -161,7 +161,13 @@ static u64 rapl_timer_ms;
 
 static inline struct rapl_pmu *cpu_to_rapl_pmu(unsigned int cpu)
 {
-	return rapl_pmus->pmus[topology_logical_package_id(cpu)];
+	unsigned int pkgid = topology_logical_package_id(cpu);
+
+	/*
+	 * The unsigned check also catches the '-1' return value for non
+	 * existent mappings in the topology map.
+	 */
+	return pkgid < rapl_pmus->maxpkg ? rapl_pmus->pmus[pkgid] : NULL;
 }
 
 static inline u64 rapl_read_counter(struct perf_event *event)
@@ -402,6 +408,8 @@ static int rapl_pmu_event_init(struct perf_event *event)
 
 	/* must be done before validate_group */
 	pmu = cpu_to_rapl_pmu(event->cpu);
+	if (!pmu)
+		return -EINVAL;
 	event->cpu = pmu->cpu;
 	event->pmu_private = pmu;
 	event->hw.event_base = msr;
@@ -585,6 +593,20 @@ static int rapl_cpu_online(unsigned int cpu)
 	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
 	int target;
 
+	if (!pmu) {
+		pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
+		if (!pmu)
+			return -ENOMEM;
+
+		raw_spin_lock_init(&pmu->lock);
+		INIT_LIST_HEAD(&pmu->active_list);
+		pmu->pmu = &rapl_pmus->pmu;
+		pmu->timer_interval = ms_to_ktime(rapl_timer_ms);
+		rapl_hrtimer_init(pmu);
+
+		rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu;
+	}
+
 	/*
 	 * Check if there is an online cpu in the package which collects rapl
 	 * events already.
@@ -598,27 +620,6 @@ static int rapl_cpu_online(unsigned int cpu)
 	return 0;
 }
 
-static int rapl_cpu_prepare(unsigned int cpu)
-{
-	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
-
-	if (pmu)
-		return 0;
-
-	pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
-	if (!pmu)
-		return -ENOMEM;
-
-	raw_spin_lock_init(&pmu->lock);
-	INIT_LIST_HEAD(&pmu->active_list);
-	pmu->pmu = &rapl_pmus->pmu;
-	pmu->timer_interval = ms_to_ktime(rapl_timer_ms);
-	pmu->cpu = -1;
-	rapl_hrtimer_init(pmu);
-	rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu;
-	return 0;
-}
-
 static int rapl_check_hw_unit(bool apply_quirk)
 {
 	u64 msr_rapl_power_unit_bits;
@@ -803,29 +804,21 @@ static int __init rapl_pmu_init(void)
 	/*
 	 * Install callbacks. Core will call them for each online cpu.
 	 */
-
-	ret = cpuhp_setup_state(CPUHP_PERF_X86_RAPL_PREP, "perf/x86/rapl:prepare",
-				rapl_cpu_prepare, NULL);
-	if (ret)
-		goto out;
-
 	ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_RAPL_ONLINE,
 				"perf/x86/rapl:online",
 				rapl_cpu_online, rapl_cpu_offline);
 	if (ret)
-		goto out1;
+		goto out;
 
 	ret = perf_pmu_register(&rapl_pmus->pmu, "power", -1);
 	if (ret)
-		goto out2;
+		goto out1;
 
 	rapl_advertise();
 	return 0;
 
-out2:
-	cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE);
 out1:
-	cpuhp_remove_state(CPUHP_PERF_X86_RAPL_PREP);
+	cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE);
 out:
 	pr_warn("Initialization failed (%d), disabled\n", ret);
 	cleanup_rapl_pmus();
@@ -836,7 +829,6 @@ module_init(rapl_pmu_init);
 static void __exit intel_rapl_exit(void)
 {
 	cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_RAPL_ONLINE);
-	cpuhp_remove_state_nocalls(CPUHP_PERF_X86_RAPL_PREP);
 	perf_pmu_unregister(&rapl_pmus->pmu);
 	cleanup_rapl_pmus();
 }
diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c
index 8c4ccdc3a3f3..1ab45976474d 100644
--- a/arch/x86/events/intel/uncore.c
+++ b/arch/x86/events/intel/uncore.c
@@ -100,7 +100,13 @@ ssize_t uncore_event_show(struct kobject *kobj,
 
 struct intel_uncore_box *uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu)
 {
-	return pmu->boxes[topology_logical_package_id(cpu)];
+	unsigned int pkgid = topology_logical_package_id(cpu);
+
+	/*
+	 * The unsigned check also catches the '-1' return value for non
+	 * existent mappings in the topology map.
+	 */
+	return pkgid < max_packages ? pmu->boxes[pkgid] : NULL;
 }
 
 u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event)
@@ -764,30 +770,6 @@ static void uncore_pmu_unregister(struct intel_uncore_pmu *pmu)
 	pmu->registered = false;
 }
 
-static void __uncore_exit_boxes(struct intel_uncore_type *type, int cpu)
-{
-	struct intel_uncore_pmu *pmu = type->pmus;
-	struct intel_uncore_box *box;
-	int i, pkg;
-
-	if (pmu) {
-		pkg = topology_physical_package_id(cpu);
-		for (i = 0; i < type->num_boxes; i++, pmu++) {
-			box = pmu->boxes[pkg];
-			if (box)
-				uncore_box_exit(box);
-		}
-	}
-}
-
-static void uncore_exit_boxes(void *dummy)
-{
-	struct intel_uncore_type **types;
-
-	for (types = uncore_msr_uncores; *types; types++)
-		__uncore_exit_boxes(*types++, smp_processor_id());
-}
-
 static void uncore_free_boxes(struct intel_uncore_pmu *pmu)
 {
 	int pkg;
@@ -1058,86 +1040,6 @@ static void uncore_pci_exit(void)
 	}
 }
 
-static int uncore_cpu_dying(unsigned int cpu)
-{
-	struct intel_uncore_type *type, **types = uncore_msr_uncores;
-	struct intel_uncore_pmu *pmu;
-	struct intel_uncore_box *box;
-	int i, pkg;
-
-	pkg = topology_logical_package_id(cpu);
-	for (; *types; types++) {
-		type = *types;
-		pmu = type->pmus;
-		for (i = 0; i < type->num_boxes; i++, pmu++) {
-			box = pmu->boxes[pkg];
-			if (box && atomic_dec_return(&box->refcnt) == 0)
-				uncore_box_exit(box);
-		}
-	}
-	return 0;
-}
-
-static int first_init;
-
-static int uncore_cpu_starting(unsigned int cpu)
-{
-	struct intel_uncore_type *type, **types = uncore_msr_uncores;
-	struct intel_uncore_pmu *pmu;
-	struct intel_uncore_box *box;
-	int i, pkg, ncpus = 1;
-
-	if (first_init) {
-		/*
-		 * On init we get the number of online cpus in the package
-		 * and set refcount for all of them.
-		 */
-		ncpus = cpumask_weight(topology_core_cpumask(cpu));
-	}
-
-	pkg = topology_logical_package_id(cpu);
-	for (; *types; types++) {
-		type = *types;
-		pmu = type->pmus;
-		for (i = 0; i < type->num_boxes; i++, pmu++) {
-			box = pmu->boxes[pkg];
-			if (!box)
-				continue;
-			/* The first cpu on a package activates the box */
-			if (atomic_add_return(ncpus, &box->refcnt) == ncpus)
-				uncore_box_init(box);
-		}
-	}
-
-	return 0;
-}
-
-static int uncore_cpu_prepare(unsigned int cpu)
-{
-	struct intel_uncore_type *type, **types = uncore_msr_uncores;
-	struct intel_uncore_pmu *pmu;
-	struct intel_uncore_box *box;
-	int i, pkg;
-
-	pkg = topology_logical_package_id(cpu);
-	for (; *types; types++) {
-		type = *types;
-		pmu = type->pmus;
-		for (i = 0; i < type->num_boxes; i++, pmu++) {
-			if (pmu->boxes[pkg])
-				continue;
-			/* First cpu of a package allocates the box */
-			box = uncore_alloc_box(type, cpu_to_node(cpu));
-			if (!box)
-				return -ENOMEM;
-			box->pmu = pmu;
-			box->pkgid = pkg;
-			pmu->boxes[pkg] = box;
-		}
-	}
-	return 0;
-}
-
 static void uncore_change_type_ctx(struct intel_uncore_type *type, int old_cpu,
 				   int new_cpu)
 {
@@ -1177,12 +1079,14 @@ static void uncore_change_context(struct intel_uncore_type **uncores,
 
 static int uncore_event_cpu_offline(unsigned int cpu)
 {
-	int target;
+	struct intel_uncore_type *type, **types = uncore_msr_uncores;
+	struct intel_uncore_pmu *pmu;
+	struct intel_uncore_box *box;
+	int i, pkg, target;
 
 	/* Check if exiting cpu is used for collecting uncore events */
 	if (!cpumask_test_and_clear_cpu(cpu, &uncore_cpu_mask))
-		return 0;
-
+		goto unref;
 	/* Find a new cpu to collect uncore events */
 	target = cpumask_any_but(topology_core_cpumask(cpu), cpu);
 
@@ -1194,12 +1098,82 @@ static int uncore_event_cpu_offline(unsigned int cpu)
 
 	uncore_change_context(uncore_msr_uncores, cpu, target);
 	uncore_change_context(uncore_pci_uncores, cpu, target);
+
+unref:
+	/* Clear the references */
+	pkg = topology_logical_package_id(cpu);
+	for (; *types; types++) {
+		type = *types;
+		pmu = type->pmus;
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			box = pmu->boxes[pkg];
+			if (box && atomic_dec_return(&box->refcnt) == 0)
+				uncore_box_exit(box);
+		}
+	}
 	return 0;
 }
 
+static int allocate_boxes(struct intel_uncore_type **types,
+			 unsigned int pkg, unsigned int cpu)
+{
+	struct intel_uncore_box *box, *tmp;
+	struct intel_uncore_type *type;
+	struct intel_uncore_pmu *pmu;
+	LIST_HEAD(allocated);
+	int i;
+
+	/* Try to allocate all required boxes */
+	for (; *types; types++) {
+		type = *types;
+		pmu = type->pmus;
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			if (pmu->boxes[pkg])
+				continue;
+			box = uncore_alloc_box(type, cpu_to_node(cpu));
+			if (!box)
+				goto cleanup;
+			box->pmu = pmu;
+			box->pkgid = pkg;
+			list_add(&box->active_list, &allocated);
+		}
+	}
+	/* Install them in the pmus */
+	list_for_each_entry_safe(box, tmp, &allocated, active_list) {
+		list_del_init(&box->active_list);
+		box->pmu->boxes[pkg] = box;
+	}
+	return 0;
+
+cleanup:
+	list_for_each_entry_safe(box, tmp, &allocated, active_list) {
+		list_del_init(&box->active_list);
+		kfree(box);
+	}
+	return -ENOMEM;
+}
+
 static int uncore_event_cpu_online(unsigned int cpu)
 {
-	int target;
+	struct intel_uncore_type *type, **types = uncore_msr_uncores;
+	struct intel_uncore_pmu *pmu;
+	struct intel_uncore_box *box;
+	int i, ret, pkg, target;
+
+	pkg = topology_logical_package_id(cpu);
+	ret = allocate_boxes(types, pkg, cpu);
+	if (ret)
+		return ret;
+
+	for (; *types; types++) {
+		type = *types;
+		pmu = type->pmus;
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			box = pmu->boxes[pkg];
+			if (!box && atomic_inc_return(&box->refcnt) == 1)
+				uncore_box_init(box);
+		}
+	}
 
 	/*
 	 * Check if there is an online cpu in the package
@@ -1389,38 +1363,16 @@ static int __init intel_uncore_init(void)
 	if (cret && pret)
 		return -ENODEV;
 
-	/*
-	 * Install callbacks. Core will call them for each online cpu.
-	 *
-	 * The first online cpu of each package allocates and takes
-	 * the refcounts for all other online cpus in that package.
-	 * If msrs are not enabled no allocation is required and
-	 * uncore_cpu_prepare() is not called for each online cpu.
-	 */
-	if (!cret) {
-	       ret = cpuhp_setup_state(CPUHP_PERF_X86_UNCORE_PREP,
-				       "perf/x86/intel/uncore:prepare",
-				       uncore_cpu_prepare, NULL);
-		if (ret)
-			goto err;
-	} else {
-		cpuhp_setup_state_nocalls(CPUHP_PERF_X86_UNCORE_PREP,
-					  "perf/x86/intel/uncore:prepare",
-					  uncore_cpu_prepare, NULL);
-	}
-	first_init = 1;
-	cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_STARTING,
-			  "perf/x86/uncore:starting",
-			  uncore_cpu_starting, uncore_cpu_dying);
-	first_init = 0;
-	cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE,
-			  "perf/x86/uncore:online",
-			  uncore_event_cpu_online, uncore_event_cpu_offline);
+	/* Install hotplug callbacks to setup the targets for each package */
+	ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE,
+				"perf/x86/intel/uncore:online",
+				uncore_event_cpu_online,
+				uncore_event_cpu_offline);
+	if (ret)
+		goto err;
 	return 0;
 
 err:
-	/* Undo box->init_box() */
-	on_each_cpu_mask(&uncore_cpu_mask, uncore_exit_boxes, NULL, 1);
 	uncore_types_exit(uncore_msr_uncores);
 	uncore_pci_exit();
 	return ret;
@@ -1429,9 +1381,7 @@ module_init(intel_uncore_init);
 
 static void __exit intel_uncore_exit(void)
 {
-	cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_UNCORE_ONLINE);
-	cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_UNCORE_STARTING);
-	cpuhp_remove_state_nocalls(CPUHP_PERF_X86_UNCORE_PREP);
+	cpuhp_remove_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE);
 	uncore_types_exit(uncore_msr_uncores);
 	uncore_pci_exit();
 }
diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h
index d936a0021839..921acaaa1601 100644
--- a/include/linux/cpuhotplug.h
+++ b/include/linux/cpuhotplug.h
@@ -8,9 +8,7 @@ enum cpuhp_state {
 	CPUHP_CREATE_THREADS,
 	CPUHP_PERF_PREPARE,
 	CPUHP_PERF_X86_PREPARE,
-	CPUHP_PERF_X86_UNCORE_PREP,
 	CPUHP_PERF_X86_AMD_UNCORE_PREP,
-	CPUHP_PERF_X86_RAPL_PREP,
 	CPUHP_PERF_BFIN,
 	CPUHP_PERF_POWER,
 	CPUHP_PERF_SUPERH,
@@ -86,7 +84,6 @@ enum cpuhp_state {
 	CPUHP_AP_IRQ_ARMADA_XP_STARTING,
 	CPUHP_AP_IRQ_BCM2836_STARTING,
 	CPUHP_AP_ARM_MVEBU_COHERENCY,
-	CPUHP_AP_PERF_X86_UNCORE_STARTING,
 	CPUHP_AP_PERF_X86_AMD_UNCORE_STARTING,
 	CPUHP_AP_PERF_X86_STARTING,
 	CPUHP_AP_PERF_X86_AMD_IBS_STARTING,
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 110b38a58493..e5aaa806702d 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -1469,7 +1469,6 @@ ctx_group_list(struct perf_event *event, struct perf_event_context *ctx)
 static void
 list_add_event(struct perf_event *event, struct perf_event_context *ctx)
 {
-
 	lockdep_assert_held(&ctx->lock);
 
 	WARN_ON_ONCE(event->attach_state & PERF_ATTACH_CONTEXT);
@@ -1624,6 +1623,8 @@ static void perf_group_attach(struct perf_event *event)
 {
 	struct perf_event *group_leader = event->group_leader, *pos;
 
+	lockdep_assert_held(&event->ctx->lock);
+
 	/*
 	 * We can have double attach due to group movement in perf_event_open.
 	 */
@@ -1697,6 +1698,8 @@ static void perf_group_detach(struct perf_event *event)
 	struct perf_event *sibling, *tmp;
 	struct list_head *list = NULL;
 
+	lockdep_assert_held(&event->ctx->lock);
+
 	/*
 	 * We can have double detach due to exit/hot-unplug + close.
 	 */
@@ -1895,9 +1898,29 @@ __perf_remove_from_context(struct perf_event *event,
  */
 static void perf_remove_from_context(struct perf_event *event, unsigned long flags)
 {
-	lockdep_assert_held(&event->ctx->mutex);
+	struct perf_event_context *ctx = event->ctx;
+
+	lockdep_assert_held(&ctx->mutex);
 
 	event_function_call(event, __perf_remove_from_context, (void *)flags);
+
+	/*
+	 * The above event_function_call() can NO-OP when it hits
+	 * TASK_TOMBSTONE. In that case we must already have been detached
+	 * from the context (by perf_event_exit_event()) but the grouping
+	 * might still be in-tact.
+	 */
+	WARN_ON_ONCE(event->attach_state & PERF_ATTACH_CONTEXT);
+	if ((flags & DETACH_GROUP) &&
+	    (event->attach_state & PERF_ATTACH_GROUP)) {
+		/*
+		 * Since in that case we cannot possibly be scheduled, simply
+		 * detach now.
+		 */
+		raw_spin_lock_irq(&ctx->lock);
+		perf_group_detach(event);
+		raw_spin_unlock_irq(&ctx->lock);
+	}
 }
 
 /*
@@ -6609,6 +6632,27 @@ static void perf_event_mmap_event(struct perf_mmap_event *mmap_event)
 	char *buf = NULL;
 	char *name;
 
+	if (vma->vm_flags & VM_READ)
+		prot |= PROT_READ;
+	if (vma->vm_flags & VM_WRITE)
+		prot |= PROT_WRITE;
+	if (vma->vm_flags & VM_EXEC)
+		prot |= PROT_EXEC;
+
+	if (vma->vm_flags & VM_MAYSHARE)
+		flags = MAP_SHARED;
+	else
+		flags = MAP_PRIVATE;
+
+	if (vma->vm_flags & VM_DENYWRITE)
+		flags |= MAP_DENYWRITE;
+	if (vma->vm_flags & VM_MAYEXEC)
+		flags |= MAP_EXECUTABLE;
+	if (vma->vm_flags & VM_LOCKED)
+		flags |= MAP_LOCKED;
+	if (vma->vm_flags & VM_HUGETLB)
+		flags |= MAP_HUGETLB;
+
 	if (file) {
 		struct inode *inode;
 		dev_t dev;
@@ -6635,27 +6679,6 @@ static void perf_event_mmap_event(struct perf_mmap_event *mmap_event)
 		maj = MAJOR(dev);
 		min = MINOR(dev);
 
-		if (vma->vm_flags & VM_READ)
-			prot |= PROT_READ;
-		if (vma->vm_flags & VM_WRITE)
-			prot |= PROT_WRITE;
-		if (vma->vm_flags & VM_EXEC)
-			prot |= PROT_EXEC;
-
-		if (vma->vm_flags & VM_MAYSHARE)
-			flags = MAP_SHARED;
-		else
-			flags = MAP_PRIVATE;
-
-		if (vma->vm_flags & VM_DENYWRITE)
-			flags |= MAP_DENYWRITE;
-		if (vma->vm_flags & VM_MAYEXEC)
-			flags |= MAP_EXECUTABLE;
-		if (vma->vm_flags & VM_LOCKED)
-			flags |= MAP_LOCKED;
-		if (vma->vm_flags & VM_HUGETLB)
-			flags |= MAP_HUGETLB;
-
 		goto got_name;
 	} else {
 		if (vma->vm_ops && vma->vm_ops->name) {