Initial commit: Linux kernel 4.19 for Unplugged OS

kernel/events/Makefile

@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0
+ifdef CONFIG_FUNCTION_TRACER
+CFLAGS_REMOVE_core.o = $(CC_FLAGS_FTRACE)
+endif
+
+obj-y := core.o ring_buffer.o callchain.o
+
+obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o
+obj-$(CONFIG_UPROBES) += uprobes.o
+

kernel/events/callchain.c

@@ -0,0 +1,260 @@
+/*
+ * Performance events callchain code, extracted from core.c:
+ *
+ *  Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
+ *  Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar
+ *  Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra
+ *  Copyright  ©  2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com>
+ *
+ * For licensing details see kernel-base/COPYING
+ */
+
+#include <linux/perf_event.h>
+#include <linux/slab.h>
+#include <linux/sched/task_stack.h>
+
+#include "internal.h"
+
+struct callchain_cpus_entries {
+	struct rcu_head			rcu_head;
+	struct perf_callchain_entry	*cpu_entries[0];
+};
+
+int sysctl_perf_event_max_stack __read_mostly = PERF_MAX_STACK_DEPTH;
+int sysctl_perf_event_max_contexts_per_stack __read_mostly = PERF_MAX_CONTEXTS_PER_STACK;
+
+static inline size_t perf_callchain_entry__sizeof(void)
+{
+	return (sizeof(struct perf_callchain_entry) +
+		sizeof(__u64) * (sysctl_perf_event_max_stack +
+				 sysctl_perf_event_max_contexts_per_stack));
+}
+
+static DEFINE_PER_CPU(int, callchain_recursion[PERF_NR_CONTEXTS]);
+static atomic_t nr_callchain_events;
+static DEFINE_MUTEX(callchain_mutex);
+static struct callchain_cpus_entries *callchain_cpus_entries;
+
+
+__weak void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
+				  struct pt_regs *regs)
+{
+}
+
+__weak void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
+				struct pt_regs *regs)
+{
+}
+
+static void release_callchain_buffers_rcu(struct rcu_head *head)
+{
+	struct callchain_cpus_entries *entries;
+	int cpu;
+
+	entries = container_of(head, struct callchain_cpus_entries, rcu_head);
+
+	for_each_possible_cpu(cpu)
+		kfree(entries->cpu_entries[cpu]);
+
+	kfree(entries);
+}
+
+static void release_callchain_buffers(void)
+{
+	struct callchain_cpus_entries *entries;
+
+	entries = callchain_cpus_entries;
+	RCU_INIT_POINTER(callchain_cpus_entries, NULL);
+	call_rcu(&entries->rcu_head, release_callchain_buffers_rcu);
+}
+
+static int alloc_callchain_buffers(void)
+{
+	int cpu;
+	int size;
+	struct callchain_cpus_entries *entries;
+
+	/*
+	 * We can't use the percpu allocation API for data that can be
+	 * accessed from NMI. Use a temporary manual per cpu allocation
+	 * until that gets sorted out.
+	 */
+	size = offsetof(struct callchain_cpus_entries, cpu_entries[nr_cpu_ids]);
+
+	entries = kzalloc(size, GFP_KERNEL);
+	if (!entries)
+		return -ENOMEM;
+
+	size = perf_callchain_entry__sizeof() * PERF_NR_CONTEXTS;
+
+	for_each_possible_cpu(cpu) {
+		entries->cpu_entries[cpu] = kmalloc_node(size, GFP_KERNEL,
+							 cpu_to_node(cpu));
+		if (!entries->cpu_entries[cpu])
+			goto fail;
+	}
+
+	rcu_assign_pointer(callchain_cpus_entries, entries);
+
+	return 0;
+
+fail:
+	for_each_possible_cpu(cpu)
+		kfree(entries->cpu_entries[cpu]);
+	kfree(entries);
+
+	return -ENOMEM;
+}
+
+int get_callchain_buffers(int event_max_stack)
+{
+	int err = 0;
+	int count;
+
+	mutex_lock(&callchain_mutex);
+
+	count = atomic_inc_return(&nr_callchain_events);
+	if (WARN_ON_ONCE(count < 1)) {
+		err = -EINVAL;
+		goto exit;
+	}
+
+	/*
+	 * If requesting per event more than the global cap,
+	 * return a different error to help userspace figure
+	 * this out.
+	 *
+	 * And also do it here so that we have &callchain_mutex held.
+	 */
+	if (event_max_stack > sysctl_perf_event_max_stack) {
+		err = -EOVERFLOW;
+		goto exit;
+	}
+
+	if (count == 1)
+		err = alloc_callchain_buffers();
+exit:
+	if (err)
+		atomic_dec(&nr_callchain_events);
+
+	mutex_unlock(&callchain_mutex);
+
+	return err;
+}
+
+void put_callchain_buffers(void)
+{
+	if (atomic_dec_and_mutex_lock(&nr_callchain_events, &callchain_mutex)) {
+		release_callchain_buffers();
+		mutex_unlock(&callchain_mutex);
+	}
+}
+
+static struct perf_callchain_entry *get_callchain_entry(int *rctx)
+{
+	int cpu;
+	struct callchain_cpus_entries *entries;
+
+	*rctx = get_recursion_context(this_cpu_ptr(callchain_recursion));
+	if (*rctx == -1)
+		return NULL;
+
+	entries = rcu_dereference(callchain_cpus_entries);
+	if (!entries)
+		return NULL;
+
+	cpu = smp_processor_id();
+
+	return (((void *)entries->cpu_entries[cpu]) +
+		(*rctx * perf_callchain_entry__sizeof()));
+}
+
+static void
+put_callchain_entry(int rctx)
+{
+	put_recursion_context(this_cpu_ptr(callchain_recursion), rctx);
+}
+
+struct perf_callchain_entry *
+get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user,
+		   u32 max_stack, bool crosstask, bool add_mark)
+{
+	struct perf_callchain_entry *entry;
+	struct perf_callchain_entry_ctx ctx;
+	int rctx;
+
+	entry = get_callchain_entry(&rctx);
+	if (rctx == -1)
+		return NULL;
+
+	if (!entry)
+		goto exit_put;
+
+	ctx.entry     = entry;
+	ctx.max_stack = max_stack;
+	ctx.nr	      = entry->nr = init_nr;
+	ctx.contexts       = 0;
+	ctx.contexts_maxed = false;
+
+	if (kernel && !user_mode(regs)) {
+		if (add_mark)
+			perf_callchain_store_context(&ctx, PERF_CONTEXT_KERNEL);
+		perf_callchain_kernel(&ctx, regs);
+	}
+
+	if (user) {
+		if (!user_mode(regs)) {
+			if  (current->mm)
+				regs = task_pt_regs(current);
+			else
+				regs = NULL;
+		}
+
+		if (regs) {
+			mm_segment_t fs;
+
+			if (crosstask)
+				goto exit_put;
+
+			if (add_mark)
+				perf_callchain_store_context(&ctx, PERF_CONTEXT_USER);
+
+			fs = get_fs();
+			set_fs(USER_DS);
+			perf_callchain_user(&ctx, regs);
+			set_fs(fs);
+		}
+	}
+
+exit_put:
+	put_callchain_entry(rctx);
+
+	return entry;
+}
+
+/*
+ * Used for sysctl_perf_event_max_stack and
+ * sysctl_perf_event_max_contexts_per_stack.
+ */
+int perf_event_max_stack_handler(struct ctl_table *table, int write,
+				 void __user *buffer, size_t *lenp, loff_t *ppos)
+{
+	int *value = table->data;
+	int new_value = *value, ret;
+	struct ctl_table new_table = *table;
+
+	new_table.data = &new_value;
+	ret = proc_dointvec_minmax(&new_table, write, buffer, lenp, ppos);
+	if (ret || !write)
+		return ret;
+
+	mutex_lock(&callchain_mutex);
+	if (atomic_read(&nr_callchain_events))
+		ret = -EBUSY;
+	else
+		*value = new_value;
+
+	mutex_unlock(&callchain_mutex);
+
+	return ret;
+}

kernel/events/hw_breakpoint.c

@@ -0,0 +1,710 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright (C) 2007 Alan Stern
+ * Copyright (C) IBM Corporation, 2009
+ * Copyright (C) 2009, Frederic Weisbecker <fweisbec@gmail.com>
+ *
+ * Thanks to Ingo Molnar for his many suggestions.
+ *
+ * Authors: Alan Stern <stern@rowland.harvard.edu>
+ *          K.Prasad <prasad@linux.vnet.ibm.com>
+ *          Frederic Weisbecker <fweisbec@gmail.com>
+ */
+
+/*
+ * HW_breakpoint: a unified kernel/user-space hardware breakpoint facility,
+ * using the CPU's debug registers.
+ * This file contains the arch-independent routines.
+ */
+
+#include <linux/irqflags.h>
+#include <linux/kallsyms.h>
+#include <linux/notifier.h>
+#include <linux/kprobes.h>
+#include <linux/kdebug.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/percpu.h>
+#include <linux/sched.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/list.h>
+#include <linux/cpu.h>
+#include <linux/smp.h>
+#include <linux/bug.h>
+
+#include <linux/hw_breakpoint.h>
+/*
+ * Constraints data
+ */
+struct bp_cpuinfo {
+	/* Number of pinned cpu breakpoints in a cpu */
+	unsigned int	cpu_pinned;
+	/* tsk_pinned[n] is the number of tasks having n+1 breakpoints */
+	unsigned int	*tsk_pinned;
+	/* Number of non-pinned cpu/task breakpoints in a cpu */
+	unsigned int	flexible; /* XXX: placeholder, see fetch_this_slot() */
+};
+
+static DEFINE_PER_CPU(struct bp_cpuinfo, bp_cpuinfo[TYPE_MAX]);
+static int nr_slots[TYPE_MAX];
+
+static struct bp_cpuinfo *get_bp_info(int cpu, enum bp_type_idx type)
+{
+	return per_cpu_ptr(bp_cpuinfo + type, cpu);
+}
+
+/* Keep track of the breakpoints attached to tasks */
+static LIST_HEAD(bp_task_head);
+
+static int constraints_initialized;
+
+/* Gather the number of total pinned and un-pinned bp in a cpuset */
+struct bp_busy_slots {
+	unsigned int pinned;
+	unsigned int flexible;
+};
+
+/* Serialize accesses to the above constraints */
+static DEFINE_MUTEX(nr_bp_mutex);
+
+__weak int hw_breakpoint_weight(struct perf_event *bp)
+{
+	return 1;
+}
+
+static inline enum bp_type_idx find_slot_idx(u64 bp_type)
+{
+	if (bp_type & HW_BREAKPOINT_RW)
+		return TYPE_DATA;
+
+	return TYPE_INST;
+}
+
+/*
+ * Report the maximum number of pinned breakpoints a task
+ * have in this cpu
+ */
+static unsigned int max_task_bp_pinned(int cpu, enum bp_type_idx type)
+{
+	unsigned int *tsk_pinned = get_bp_info(cpu, type)->tsk_pinned;
+	int i;
+
+	for (i = nr_slots[type] - 1; i >= 0; i--) {
+		if (tsk_pinned[i] > 0)
+			return i + 1;
+	}
+
+	return 0;
+}
+
+/*
+ * Count the number of breakpoints of the same type and same task.
+ * The given event must be not on the list.
+ */
+static int task_bp_pinned(int cpu, struct perf_event *bp, enum bp_type_idx type)
+{
+	struct task_struct *tsk = bp->hw.target;
+	struct perf_event *iter;
+	int count = 0;
+
+	list_for_each_entry(iter, &bp_task_head, hw.bp_list) {
+		if (iter->hw.target == tsk &&
+		    find_slot_idx(iter->attr.bp_type) == type &&
+		    (iter->cpu < 0 || cpu == iter->cpu))
+			count += hw_breakpoint_weight(iter);
+	}
+
+	return count;
+}
+
+static const struct cpumask *cpumask_of_bp(struct perf_event *bp)
+{
+	if (bp->cpu >= 0)
+		return cpumask_of(bp->cpu);
+	return cpu_possible_mask;
+}
+
+/*
+ * Report the number of pinned/un-pinned breakpoints we have in
+ * a given cpu (cpu > -1) or in all of them (cpu = -1).
+ */
+static void
+fetch_bp_busy_slots(struct bp_busy_slots *slots, struct perf_event *bp,
+		    enum bp_type_idx type)
+{
+	const struct cpumask *cpumask = cpumask_of_bp(bp);
+	int cpu;
+
+	for_each_cpu(cpu, cpumask) {
+		struct bp_cpuinfo *info = get_bp_info(cpu, type);
+		int nr;
+
+		nr = info->cpu_pinned;
+		if (!bp->hw.target)
+			nr += max_task_bp_pinned(cpu, type);
+		else
+			nr += task_bp_pinned(cpu, bp, type);
+
+		if (nr > slots->pinned)
+			slots->pinned = nr;
+
+		nr = info->flexible;
+		if (nr > slots->flexible)
+			slots->flexible = nr;
+	}
+}
+
+/*
+ * For now, continue to consider flexible as pinned, until we can
+ * ensure no flexible event can ever be scheduled before a pinned event
+ * in a same cpu.
+ */
+static void
+fetch_this_slot(struct bp_busy_slots *slots, int weight)
+{
+	slots->pinned += weight;
+}
+
+/*
+ * Add a pinned breakpoint for the given task in our constraint table
+ */
+static void toggle_bp_task_slot(struct perf_event *bp, int cpu,
+				enum bp_type_idx type, int weight)
+{
+	unsigned int *tsk_pinned = get_bp_info(cpu, type)->tsk_pinned;
+	int old_idx, new_idx;
+
+	old_idx = task_bp_pinned(cpu, bp, type) - 1;
+	new_idx = old_idx + weight;
+
+	if (old_idx >= 0)
+		tsk_pinned[old_idx]--;
+	if (new_idx >= 0)
+		tsk_pinned[new_idx]++;
+}
+
+/*
+ * Add/remove the given breakpoint in our constraint table
+ */
+static void
+toggle_bp_slot(struct perf_event *bp, bool enable, enum bp_type_idx type,
+	       int weight)
+{
+	const struct cpumask *cpumask = cpumask_of_bp(bp);
+	int cpu;
+
+	if (!enable)
+		weight = -weight;
+
+	/* Pinned counter cpu profiling */
+	if (!bp->hw.target) {
+		get_bp_info(bp->cpu, type)->cpu_pinned += weight;
+		return;
+	}
+
+	/* Pinned counter task profiling */
+	for_each_cpu(cpu, cpumask)
+		toggle_bp_task_slot(bp, cpu, type, weight);
+
+	if (enable)
+		list_add_tail(&bp->hw.bp_list, &bp_task_head);
+	else
+		list_del(&bp->hw.bp_list);
+}
+
+/*
+ * Function to perform processor-specific cleanup during unregistration
+ */
+__weak void arch_unregister_hw_breakpoint(struct perf_event *bp)
+{
+	/*
+	 * A weak stub function here for those archs that don't define
+	 * it inside arch/.../kernel/hw_breakpoint.c
+	 */
+}
+
+/*
+ * Contraints to check before allowing this new breakpoint counter:
+ *
+ *  == Non-pinned counter == (Considered as pinned for now)
+ *
+ *   - If attached to a single cpu, check:
+ *
+ *       (per_cpu(info->flexible, cpu) || (per_cpu(info->cpu_pinned, cpu)
+ *           + max(per_cpu(info->tsk_pinned, cpu)))) < HBP_NUM
+ *
+ *       -> If there are already non-pinned counters in this cpu, it means
+ *          there is already a free slot for them.
+ *          Otherwise, we check that the maximum number of per task
+ *          breakpoints (for this cpu) plus the number of per cpu breakpoint
+ *          (for this cpu) doesn't cover every registers.
+ *
+ *   - If attached to every cpus, check:
+ *
+ *       (per_cpu(info->flexible, *) || (max(per_cpu(info->cpu_pinned, *))
+ *           + max(per_cpu(info->tsk_pinned, *)))) < HBP_NUM
+ *
+ *       -> This is roughly the same, except we check the number of per cpu
+ *          bp for every cpu and we keep the max one. Same for the per tasks
+ *          breakpoints.
+ *
+ *
+ * == Pinned counter ==
+ *
+ *   - If attached to a single cpu, check:
+ *
+ *       ((per_cpu(info->flexible, cpu) > 1) + per_cpu(info->cpu_pinned, cpu)
+ *            + max(per_cpu(info->tsk_pinned, cpu))) < HBP_NUM
+ *
+ *       -> Same checks as before. But now the info->flexible, if any, must keep
+ *          one register at least (or they will never be fed).
+ *
+ *   - If attached to every cpus, check:
+ *
+ *       ((per_cpu(info->flexible, *) > 1) + max(per_cpu(info->cpu_pinned, *))
+ *            + max(per_cpu(info->tsk_pinned, *))) < HBP_NUM
+ */
+static int __reserve_bp_slot(struct perf_event *bp, u64 bp_type)
+{
+	struct bp_busy_slots slots = {0};
+	enum bp_type_idx type;
+	int weight;
+
+	/* We couldn't initialize breakpoint constraints on boot */
+	if (!constraints_initialized)
+		return -ENOMEM;
+
+	/* Basic checks */
+	if (bp_type == HW_BREAKPOINT_EMPTY ||
+	    bp_type == HW_BREAKPOINT_INVALID)
+		return -EINVAL;
+
+	type = find_slot_idx(bp_type);
+	weight = hw_breakpoint_weight(bp);
+
+	fetch_bp_busy_slots(&slots, bp, type);
+	/*
+	 * Simulate the addition of this breakpoint to the constraints
+	 * and see the result.
+	 */
+	fetch_this_slot(&slots, weight);
+
+	/* Flexible counters need to keep at least one slot */
+	if (slots.pinned + (!!slots.flexible) > nr_slots[type])
+		return -ENOSPC;
+
+	toggle_bp_slot(bp, true, type, weight);
+
+	return 0;
+}
+
+int reserve_bp_slot(struct perf_event *bp)
+{
+	int ret;
+
+	mutex_lock(&nr_bp_mutex);
+
+	ret = __reserve_bp_slot(bp, bp->attr.bp_type);
+
+	mutex_unlock(&nr_bp_mutex);
+
+	return ret;
+}
+
+static void __release_bp_slot(struct perf_event *bp, u64 bp_type)
+{
+	enum bp_type_idx type;
+	int weight;
+
+	type = find_slot_idx(bp_type);
+	weight = hw_breakpoint_weight(bp);
+	toggle_bp_slot(bp, false, type, weight);
+}
+
+void release_bp_slot(struct perf_event *bp)
+{
+	mutex_lock(&nr_bp_mutex);
+
+	arch_unregister_hw_breakpoint(bp);
+	__release_bp_slot(bp, bp->attr.bp_type);
+
+	mutex_unlock(&nr_bp_mutex);
+}
+
+static int __modify_bp_slot(struct perf_event *bp, u64 old_type, u64 new_type)
+{
+	int err;
+
+	__release_bp_slot(bp, old_type);
+
+	err = __reserve_bp_slot(bp, new_type);
+	if (err) {
+		/*
+		 * Reserve the old_type slot back in case
+		 * there's no space for the new type.
+		 *
+		 * This must succeed, because we just released
+		 * the old_type slot in the __release_bp_slot
+		 * call above. If not, something is broken.
+		 */
+		WARN_ON(__reserve_bp_slot(bp, old_type));
+	}
+
+	return err;
+}
+
+static int modify_bp_slot(struct perf_event *bp, u64 old_type, u64 new_type)
+{
+	int ret;
+
+	mutex_lock(&nr_bp_mutex);
+	ret = __modify_bp_slot(bp, old_type, new_type);
+	mutex_unlock(&nr_bp_mutex);
+	return ret;
+}
+
+/*
+ * Allow the kernel debugger to reserve breakpoint slots without
+ * taking a lock using the dbg_* variant of for the reserve and
+ * release breakpoint slots.
+ */
+int dbg_reserve_bp_slot(struct perf_event *bp)
+{
+	if (mutex_is_locked(&nr_bp_mutex))
+		return -1;
+
+	return __reserve_bp_slot(bp, bp->attr.bp_type);
+}
+
+int dbg_release_bp_slot(struct perf_event *bp)
+{
+	if (mutex_is_locked(&nr_bp_mutex))
+		return -1;
+
+	__release_bp_slot(bp, bp->attr.bp_type);
+
+	return 0;
+}
+
+static int hw_breakpoint_parse(struct perf_event *bp,
+			       const struct perf_event_attr *attr,
+			       struct arch_hw_breakpoint *hw)
+{
+	int err;
+
+	err = hw_breakpoint_arch_parse(bp, attr, hw);
+	if (err)
+		return err;
+
+	if (arch_check_bp_in_kernelspace(hw)) {
+		if (attr->exclude_kernel)
+			return -EINVAL;
+		/*
+		 * Don't let unprivileged users set a breakpoint in the trap
+		 * path to avoid trap recursion attacks.
+		 */
+		if (!capable(CAP_SYS_ADMIN))
+			return -EPERM;
+	}
+
+	return 0;
+}
+
+int register_perf_hw_breakpoint(struct perf_event *bp)
+{
+	struct arch_hw_breakpoint hw = { };
+	int err;
+
+	err = reserve_bp_slot(bp);
+	if (err)
+		return err;
+
+	err = hw_breakpoint_parse(bp, &bp->attr, &hw);
+	if (err) {
+		release_bp_slot(bp);
+		return err;
+	}
+
+	bp->hw.info = hw;
+
+	return 0;
+}
+
+/**
+ * register_user_hw_breakpoint - register a hardware breakpoint for user space
+ * @attr: breakpoint attributes
+ * @triggered: callback to trigger when we hit the breakpoint
+ * @tsk: pointer to 'task_struct' of the process to which the address belongs
+ */
+struct perf_event *
+register_user_hw_breakpoint(struct perf_event_attr *attr,
+			    perf_overflow_handler_t triggered,
+			    void *context,
+			    struct task_struct *tsk)
+{
+	return perf_event_create_kernel_counter(attr, -1, tsk, triggered,
+						context);
+}
+EXPORT_SYMBOL_GPL(register_user_hw_breakpoint);
+
+static void hw_breakpoint_copy_attr(struct perf_event_attr *to,
+				    struct perf_event_attr *from)
+{
+	to->bp_addr = from->bp_addr;
+	to->bp_type = from->bp_type;
+	to->bp_len  = from->bp_len;
+	to->disabled = from->disabled;
+}
+
+int
+modify_user_hw_breakpoint_check(struct perf_event *bp, struct perf_event_attr *attr,
+			        bool check)
+{
+	struct arch_hw_breakpoint hw = { };
+	int err;
+
+	err = hw_breakpoint_parse(bp, attr, &hw);
+	if (err)
+		return err;
+
+	if (check) {
+		struct perf_event_attr old_attr;
+
+		old_attr = bp->attr;
+		hw_breakpoint_copy_attr(&old_attr, attr);
+		if (memcmp(&old_attr, attr, sizeof(*attr)))
+			return -EINVAL;
+	}
+
+	if (bp->attr.bp_type != attr->bp_type) {
+		err = modify_bp_slot(bp, bp->attr.bp_type, attr->bp_type);
+		if (err)
+			return err;
+	}
+
+	hw_breakpoint_copy_attr(&bp->attr, attr);
+	bp->hw.info = hw;
+
+	return 0;
+}
+
+/**
+ * modify_user_hw_breakpoint - modify a user-space hardware breakpoint
+ * @bp: the breakpoint structure to modify
+ * @attr: new breakpoint attributes
+ */
+int modify_user_hw_breakpoint(struct perf_event *bp, struct perf_event_attr *attr)
+{
+	int err;
+
+	/*
+	 * modify_user_hw_breakpoint can be invoked with IRQs disabled and hence it
+	 * will not be possible to raise IPIs that invoke __perf_event_disable.
+	 * So call the function directly after making sure we are targeting the
+	 * current task.
+	 */
+	if (irqs_disabled() && bp->ctx && bp->ctx->task == current)
+		perf_event_disable_local(bp);
+	else
+		perf_event_disable(bp);
+
+	err = modify_user_hw_breakpoint_check(bp, attr, false);
+
+	if (!bp->attr.disabled)
+		perf_event_enable(bp);
+
+	return err;
+}
+EXPORT_SYMBOL_GPL(modify_user_hw_breakpoint);
+
+/**
+ * unregister_hw_breakpoint - unregister a user-space hardware breakpoint
+ * @bp: the breakpoint structure to unregister
+ */
+void unregister_hw_breakpoint(struct perf_event *bp)
+{
+	if (!bp)
+		return;
+	perf_event_release_kernel(bp);
+}
+EXPORT_SYMBOL_GPL(unregister_hw_breakpoint);
+
+/**
+ * register_wide_hw_breakpoint - register a wide breakpoint in the kernel
+ * @attr: breakpoint attributes
+ * @triggered: callback to trigger when we hit the breakpoint
+ *
+ * @return a set of per_cpu pointers to perf events
+ */
+struct perf_event * __percpu *
+register_wide_hw_breakpoint(struct perf_event_attr *attr,
+			    perf_overflow_handler_t triggered,
+			    void *context)
+{
+	struct perf_event * __percpu *cpu_events, *bp;
+	long err = 0;
+	int cpu;
+
+	cpu_events = alloc_percpu(typeof(*cpu_events));
+	if (!cpu_events)
+		return (void __percpu __force *)ERR_PTR(-ENOMEM);
+
+	get_online_cpus();
+	for_each_online_cpu(cpu) {
+		bp = perf_event_create_kernel_counter(attr, cpu, NULL,
+						      triggered, context);
+		if (IS_ERR(bp)) {
+			err = PTR_ERR(bp);
+			break;
+		}
+
+		per_cpu(*cpu_events, cpu) = bp;
+	}
+	put_online_cpus();
+
+	if (likely(!err))
+		return cpu_events;
+
+	unregister_wide_hw_breakpoint(cpu_events);
+	return (void __percpu __force *)ERR_PTR(err);
+}
+EXPORT_SYMBOL_GPL(register_wide_hw_breakpoint);
+
+/**
+ * unregister_wide_hw_breakpoint - unregister a wide breakpoint in the kernel
+ * @cpu_events: the per cpu set of events to unregister
+ */
+void unregister_wide_hw_breakpoint(struct perf_event * __percpu *cpu_events)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu)
+		unregister_hw_breakpoint(per_cpu(*cpu_events, cpu));
+
+	free_percpu(cpu_events);
+}
+EXPORT_SYMBOL_GPL(unregister_wide_hw_breakpoint);
+
+static struct notifier_block hw_breakpoint_exceptions_nb = {
+	.notifier_call = hw_breakpoint_exceptions_notify,
+	/* we need to be notified first */
+	.priority = 0x7fffffff
+};
+
+static void bp_perf_event_destroy(struct perf_event *event)
+{
+	release_bp_slot(event);
+}
+
+static int hw_breakpoint_event_init(struct perf_event *bp)
+{
+	int err;
+
+	if (bp->attr.type != PERF_TYPE_BREAKPOINT)
+		return -ENOENT;
+
+	/*
+	 * no branch sampling for breakpoint events
+	 */
+	if (has_branch_stack(bp))
+		return -EOPNOTSUPP;
+
+	err = register_perf_hw_breakpoint(bp);
+	if (err)
+		return err;
+
+	bp->destroy = bp_perf_event_destroy;
+
+	return 0;
+}
+
+static int hw_breakpoint_add(struct perf_event *bp, int flags)
+{
+	if (!(flags & PERF_EF_START))
+		bp->hw.state = PERF_HES_STOPPED;
+
+	if (is_sampling_event(bp)) {
+		bp->hw.last_period = bp->hw.sample_period;
+		perf_swevent_set_period(bp);
+	}
+
+	return arch_install_hw_breakpoint(bp);
+}
+
+static void hw_breakpoint_del(struct perf_event *bp, int flags)
+{
+	arch_uninstall_hw_breakpoint(bp);
+}
+
+static void hw_breakpoint_start(struct perf_event *bp, int flags)
+{
+	bp->hw.state = 0;
+}
+
+static void hw_breakpoint_stop(struct perf_event *bp, int flags)
+{
+	bp->hw.state = PERF_HES_STOPPED;
+}
+
+static struct pmu perf_breakpoint = {
+	.task_ctx_nr	= perf_sw_context, /* could eventually get its own */
+
+	.event_init	= hw_breakpoint_event_init,
+	.add		= hw_breakpoint_add,
+	.del		= hw_breakpoint_del,
+	.start		= hw_breakpoint_start,
+	.stop		= hw_breakpoint_stop,
+	.read		= hw_breakpoint_pmu_read,
+};
+
+int __init init_hw_breakpoint(void)
+{
+	int cpu, err_cpu;
+	int i;
+
+	for (i = 0; i < TYPE_MAX; i++)
+		nr_slots[i] = hw_breakpoint_slots(i);
+
+	for_each_possible_cpu(cpu) {
+		for (i = 0; i < TYPE_MAX; i++) {
+			struct bp_cpuinfo *info = get_bp_info(cpu, i);
+
+			info->tsk_pinned = kcalloc(nr_slots[i], sizeof(int),
+							GFP_KERNEL);
+			if (!info->tsk_pinned)
+				goto err_alloc;
+		}
+	}
+
+	constraints_initialized = 1;
+
+	perf_pmu_register(&perf_breakpoint, "breakpoint", PERF_TYPE_BREAKPOINT);
+
+	return register_die_notifier(&hw_breakpoint_exceptions_nb);
+
+ err_alloc:
+	for_each_possible_cpu(err_cpu) {
+		for (i = 0; i < TYPE_MAX; i++)
+			kfree(get_bp_info(err_cpu, i)->tsk_pinned);
+		if (err_cpu == cpu)
+			break;
+	}
+
+	return -ENOMEM;
+}
+
+

kernel/events/internal.h

@@ -0,0 +1,248 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _KERNEL_EVENTS_INTERNAL_H
+#define _KERNEL_EVENTS_INTERNAL_H
+
+#include <linux/hardirq.h>
+#include <linux/uaccess.h>
+
+/* Buffer handling */
+
+#define RING_BUFFER_WRITABLE		0x01
+
+struct ring_buffer {
+	atomic_t			refcount;
+	struct rcu_head			rcu_head;
+#ifdef CONFIG_PERF_USE_VMALLOC
+	struct work_struct		work;
+	int				page_order;	/* allocation order  */
+#endif
+	int				nr_pages;	/* nr of data pages  */
+	int				overwrite;	/* can overwrite itself */
+	int				paused;		/* can write into ring buffer */
+
+	atomic_t			poll;		/* POLL_ for wakeups */
+
+	local_t				head;		/* write position    */
+	local_t				nest;		/* nested writers    */
+	local_t				events;		/* event limit       */
+	local_t				wakeup;		/* wakeup stamp      */
+	local_t				lost;		/* nr records lost   */
+
+	long				watermark;	/* wakeup watermark  */
+	long				aux_watermark;
+	/* poll crap */
+	spinlock_t			event_lock;
+	struct list_head		event_list;
+
+	atomic_t			mmap_count;
+	unsigned long			mmap_locked;
+	struct user_struct		*mmap_user;
+
+	/* AUX area */
+	long				aux_head;
+	local_t				aux_nest;
+	long				aux_wakeup;	/* last aux_watermark boundary crossed by aux_head */
+	unsigned long			aux_pgoff;
+	int				aux_nr_pages;
+	int				aux_overwrite;
+	atomic_t			aux_mmap_count;
+	unsigned long			aux_mmap_locked;
+	void				(*free_aux)(void *);
+	atomic_t			aux_refcount;
+	void				**aux_pages;
+	void				*aux_priv;
+
+	struct perf_event_mmap_page	*user_page;
+	void				*data_pages[0];
+};
+
+extern void rb_free(struct ring_buffer *rb);
+
+static inline void rb_free_rcu(struct rcu_head *rcu_head)
+{
+	struct ring_buffer *rb;
+
+	rb = container_of(rcu_head, struct ring_buffer, rcu_head);
+	rb_free(rb);
+}
+
+static inline void rb_toggle_paused(struct ring_buffer *rb, bool pause)
+{
+	if (!pause && rb->nr_pages)
+		rb->paused = 0;
+	else
+		rb->paused = 1;
+}
+
+extern struct ring_buffer *
+rb_alloc(int nr_pages, long watermark, int cpu, int flags);
+extern void perf_event_wakeup(struct perf_event *event);
+extern int rb_alloc_aux(struct ring_buffer *rb, struct perf_event *event,
+			pgoff_t pgoff, int nr_pages, long watermark, int flags);
+extern void rb_free_aux(struct ring_buffer *rb);
+extern struct ring_buffer *ring_buffer_get(struct perf_event *event);
+extern void ring_buffer_put(struct ring_buffer *rb);
+
+static inline bool rb_has_aux(struct ring_buffer *rb)
+{
+	return !!rb->aux_nr_pages;
+}
+
+void perf_event_aux_event(struct perf_event *event, unsigned long head,
+			  unsigned long size, u64 flags);
+
+extern struct page *
+perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff);
+
+#ifdef CONFIG_PERF_USE_VMALLOC
+/*
+ * Back perf_mmap() with vmalloc memory.
+ *
+ * Required for architectures that have d-cache aliasing issues.
+ */
+
+static inline int page_order(struct ring_buffer *rb)
+{
+	return rb->page_order;
+}
+
+#else
+
+static inline int page_order(struct ring_buffer *rb)
+{
+	return 0;
+}
+#endif
+
+static inline unsigned long perf_data_size(struct ring_buffer *rb)
+{
+	return rb->nr_pages << (PAGE_SHIFT + page_order(rb));
+}
+
+static inline unsigned long perf_aux_size(struct ring_buffer *rb)
+{
+	return rb->aux_nr_pages << PAGE_SHIFT;
+}
+
+#define __DEFINE_OUTPUT_COPY_BODY(advance_buf, memcpy_func, ...)	\
+{									\
+	unsigned long size, written;					\
+									\
+	do {								\
+		size    = min(handle->size, len);			\
+		written = memcpy_func(__VA_ARGS__);			\
+		written = size - written;				\
+									\
+		len -= written;						\
+		handle->addr += written;				\
+		if (advance_buf)					\
+			buf += written;					\
+		handle->size -= written;				\
+		if (!handle->size) {					\
+			struct ring_buffer *rb = handle->rb;		\
+									\
+			handle->page++;					\
+			handle->page &= rb->nr_pages - 1;		\
+			handle->addr = rb->data_pages[handle->page];	\
+			handle->size = PAGE_SIZE << page_order(rb);	\
+		}							\
+	} while (len && written == size);				\
+									\
+	return len;							\
+}
+
+#define DEFINE_OUTPUT_COPY(func_name, memcpy_func)			\
+static inline unsigned long						\
+func_name(struct perf_output_handle *handle,				\
+	  const void *buf, unsigned long len)				\
+__DEFINE_OUTPUT_COPY_BODY(true, memcpy_func, handle->addr, buf, size)
+
+static inline unsigned long
+__output_custom(struct perf_output_handle *handle, perf_copy_f copy_func,
+		const void *buf, unsigned long len)
+{
+	unsigned long orig_len = len;
+	__DEFINE_OUTPUT_COPY_BODY(false, copy_func, handle->addr, buf,
+				  orig_len - len, size)
+}
+
+static inline unsigned long
+memcpy_common(void *dst, const void *src, unsigned long n)
+{
+	memcpy(dst, src, n);
+	return 0;
+}
+
+DEFINE_OUTPUT_COPY(__output_copy, memcpy_common)
+
+static inline unsigned long
+memcpy_skip(void *dst, const void *src, unsigned long n)
+{
+	return 0;
+}
+
+DEFINE_OUTPUT_COPY(__output_skip, memcpy_skip)
+
+#ifndef arch_perf_out_copy_user
+#define arch_perf_out_copy_user arch_perf_out_copy_user
+
+static inline unsigned long
+arch_perf_out_copy_user(void *dst, const void *src, unsigned long n)
+{
+	unsigned long ret;
+
+	pagefault_disable();
+	ret = __copy_from_user_inatomic(dst, src, n);
+	pagefault_enable();
+
+	return ret;
+}
+#endif
+
+DEFINE_OUTPUT_COPY(__output_copy_user, arch_perf_out_copy_user)
+
+static inline int get_recursion_context(int *recursion)
+{
+	int rctx;
+
+	if (unlikely(in_nmi()))
+		rctx = 3;
+	else if (in_irq())
+		rctx = 2;
+	else if (in_serving_softirq())
+		rctx = 1;
+	else
+		rctx = 0;
+
+	if (recursion[rctx])
+		return -1;
+
+	recursion[rctx]++;
+	barrier();
+
+	return rctx;
+}
+
+static inline void put_recursion_context(int *recursion, int rctx)
+{
+	barrier();
+	recursion[rctx]--;
+}
+
+#ifdef CONFIG_HAVE_PERF_USER_STACK_DUMP
+static inline bool arch_perf_have_user_stack_dump(void)
+{
+	return true;
+}
+
+#define perf_user_stack_pointer(regs) user_stack_pointer(regs)
+#else
+static inline bool arch_perf_have_user_stack_dump(void)
+{
+	return false;
+}
+
+#define perf_user_stack_pointer(regs) 0
+#endif /* CONFIG_HAVE_PERF_USER_STACK_DUMP */
+
+#endif /* _KERNEL_EVENTS_INTERNAL_H */

kernel/events/ring_buffer.c

@@ -0,0 +1,907 @@
+/*
+ * Performance events ring-buffer code:
+ *
+ *  Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
+ *  Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar
+ *  Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra
+ *  Copyright  ©  2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com>
+ *
+ * For licensing details see kernel-base/COPYING
+ */
+
+#include <linux/perf_event.h>
+#include <linux/vmalloc.h>
+#include <linux/slab.h>
+#include <linux/circ_buf.h>
+#include <linux/poll.h>
+#include <linux/nospec.h>
+
+#include "internal.h"
+
+static void perf_output_wakeup(struct perf_output_handle *handle)
+{
+	atomic_set(&handle->rb->poll, EPOLLIN);
+
+	handle->event->pending_wakeup = 1;
+	irq_work_queue(&handle->event->pending);
+}
+
+/*
+ * We need to ensure a later event_id doesn't publish a head when a former
+ * event isn't done writing. However since we need to deal with NMIs we
+ * cannot fully serialize things.
+ *
+ * We only publish the head (and generate a wakeup) when the outer-most
+ * event completes.
+ */
+static void perf_output_get_handle(struct perf_output_handle *handle)
+{
+	struct ring_buffer *rb = handle->rb;
+
+	preempt_disable();
+	local_inc(&rb->nest);
+	handle->wakeup = local_read(&rb->wakeup);
+}
+
+static void perf_output_put_handle(struct perf_output_handle *handle)
+{
+	struct ring_buffer *rb = handle->rb;
+	unsigned long head;
+
+again:
+	/*
+	 * In order to avoid publishing a head value that goes backwards,
+	 * we must ensure the load of @rb->head happens after we've
+	 * incremented @rb->nest.
+	 *
+	 * Otherwise we can observe a @rb->head value before one published
+	 * by an IRQ/NMI happening between the load and the increment.
+	 */
+	barrier();
+	head = local_read(&rb->head);
+
+	/*
+	 * IRQ/NMI can happen here and advance @rb->head, causing our
+	 * load above to be stale.
+	 */
+
+	/*
+	 * If this isn't the outermost nesting, we don't have to update
+	 * @rb->user_page->data_head.
+	 */
+	if (local_read(&rb->nest) > 1) {
+		local_dec(&rb->nest);
+		goto out;
+	}
+
+	/*
+	 * Since the mmap() consumer (userspace) can run on a different CPU:
+	 *
+	 *   kernel				user
+	 *
+	 *   if (LOAD ->data_tail) {		LOAD ->data_head
+	 *			(A)		smp_rmb()	(C)
+	 *	STORE $data			LOAD $data
+	 *	smp_wmb()	(B)		smp_mb()	(D)
+	 *	STORE ->data_head		STORE ->data_tail
+	 *   }
+	 *
+	 * Where A pairs with D, and B pairs with C.
+	 *
+	 * In our case (A) is a control dependency that separates the load of
+	 * the ->data_tail and the stores of $data. In case ->data_tail
+	 * indicates there is no room in the buffer to store $data we do not.
+	 *
+	 * D needs to be a full barrier since it separates the data READ
+	 * from the tail WRITE.
+	 *
+	 * For B a WMB is sufficient since it separates two WRITEs, and for C
+	 * an RMB is sufficient since it separates two READs.
+	 *
+	 * See perf_output_begin().
+	 */
+	smp_wmb(); /* B, matches C */
+	WRITE_ONCE(rb->user_page->data_head, head);
+
+	/*
+	 * We must publish the head before decrementing the nest count,
+	 * otherwise an IRQ/NMI can publish a more recent head value and our
+	 * write will (temporarily) publish a stale value.
+	 */
+	barrier();
+	local_set(&rb->nest, 0);
+
+	/*
+	 * Ensure we decrement @rb->nest before we validate the @rb->head.
+	 * Otherwise we cannot be sure we caught the 'last' nested update.
+	 */
+	barrier();
+	if (unlikely(head != local_read(&rb->head))) {
+		local_inc(&rb->nest);
+		goto again;
+	}
+
+	if (handle->wakeup != local_read(&rb->wakeup))
+		perf_output_wakeup(handle);
+
+out:
+	preempt_enable();
+}
+
+static __always_inline bool
+ring_buffer_has_space(unsigned long head, unsigned long tail,
+		      unsigned long data_size, unsigned int size,
+		      bool backward)
+{
+	if (!backward)
+		return CIRC_SPACE(head, tail, data_size) >= size;
+	else
+		return CIRC_SPACE(tail, head, data_size) >= size;
+}
+
+static __always_inline int
+__perf_output_begin(struct perf_output_handle *handle,
+		    struct perf_event *event, unsigned int size,
+		    bool backward)
+{
+	struct ring_buffer *rb;
+	unsigned long tail, offset, head;
+	int have_lost, page_shift;
+	struct {
+		struct perf_event_header header;
+		u64			 id;
+		u64			 lost;
+	} lost_event;
+
+	rcu_read_lock();
+	/*
+	 * For inherited events we send all the output towards the parent.
+	 */
+	if (event->parent)
+		event = event->parent;
+
+	rb = rcu_dereference(event->rb);
+	if (unlikely(!rb))
+		goto out;
+
+	if (unlikely(rb->paused)) {
+		if (rb->nr_pages)
+			local_inc(&rb->lost);
+		goto out;
+	}
+
+	handle->rb    = rb;
+	handle->event = event;
+
+	have_lost = local_read(&rb->lost);
+	if (unlikely(have_lost)) {
+		size += sizeof(lost_event);
+		if (event->attr.sample_id_all)
+			size += event->id_header_size;
+	}
+
+	perf_output_get_handle(handle);
+
+	do {
+		tail = READ_ONCE(rb->user_page->data_tail);
+		offset = head = local_read(&rb->head);
+		if (!rb->overwrite) {
+			if (unlikely(!ring_buffer_has_space(head, tail,
+							    perf_data_size(rb),
+							    size, backward)))
+				goto fail;
+		}
+
+		/*
+		 * The above forms a control dependency barrier separating the
+		 * @tail load above from the data stores below. Since the @tail
+		 * load is required to compute the branch to fail below.
+		 *
+		 * A, matches D; the full memory barrier userspace SHOULD issue
+		 * after reading the data and before storing the new tail
+		 * position.
+		 *
+		 * See perf_output_put_handle().
+		 */
+
+		if (!backward)
+			head += size;
+		else
+			head -= size;
+	} while (local_cmpxchg(&rb->head, offset, head) != offset);
+
+	if (backward) {
+		offset = head;
+		head = (u64)(-head);
+	}
+
+	/*
+	 * We rely on the implied barrier() by local_cmpxchg() to ensure
+	 * none of the data stores below can be lifted up by the compiler.
+	 */
+
+	if (unlikely(head - local_read(&rb->wakeup) > rb->watermark))
+		local_add(rb->watermark, &rb->wakeup);
+
+	page_shift = PAGE_SHIFT + page_order(rb);
+
+	handle->page = (offset >> page_shift) & (rb->nr_pages - 1);
+	offset &= (1UL << page_shift) - 1;
+	handle->addr = rb->data_pages[handle->page] + offset;
+	handle->size = (1UL << page_shift) - offset;
+
+	if (unlikely(have_lost)) {
+		struct perf_sample_data sample_data;
+
+		lost_event.header.size = sizeof(lost_event);
+		lost_event.header.type = PERF_RECORD_LOST;
+		lost_event.header.misc = 0;
+		lost_event.id          = event->id;
+		lost_event.lost        = local_xchg(&rb->lost, 0);
+
+		perf_event_header__init_id(&lost_event.header,
+					   &sample_data, event);
+		perf_output_put(handle, lost_event);
+		perf_event__output_id_sample(event, handle, &sample_data);
+	}
+
+	return 0;
+
+fail:
+	local_inc(&rb->lost);
+	perf_output_put_handle(handle);
+out:
+	rcu_read_unlock();
+
+	return -ENOSPC;
+}
+
+int perf_output_begin_forward(struct perf_output_handle *handle,
+			     struct perf_event *event, unsigned int size)
+{
+	return __perf_output_begin(handle, event, size, false);
+}
+
+int perf_output_begin_backward(struct perf_output_handle *handle,
+			       struct perf_event *event, unsigned int size)
+{
+	return __perf_output_begin(handle, event, size, true);
+}
+
+int perf_output_begin(struct perf_output_handle *handle,
+		      struct perf_event *event, unsigned int size)
+{
+
+	return __perf_output_begin(handle, event, size,
+				   unlikely(is_write_backward(event)));
+}
+
+unsigned int perf_output_copy(struct perf_output_handle *handle,
+		      const void *buf, unsigned int len)
+{
+	return __output_copy(handle, buf, len);
+}
+
+unsigned int perf_output_skip(struct perf_output_handle *handle,
+			      unsigned int len)
+{
+	return __output_skip(handle, NULL, len);
+}
+
+void perf_output_end(struct perf_output_handle *handle)
+{
+	perf_output_put_handle(handle);
+	rcu_read_unlock();
+}
+
+static void
+ring_buffer_init(struct ring_buffer *rb, long watermark, int flags)
+{
+	long max_size = perf_data_size(rb);
+
+	if (watermark)
+		rb->watermark = min(max_size, watermark);
+
+	if (!rb->watermark)
+		rb->watermark = max_size / 2;
+
+	if (flags & RING_BUFFER_WRITABLE)
+		rb->overwrite = 0;
+	else
+		rb->overwrite = 1;
+
+	atomic_set(&rb->refcount, 1);
+
+	INIT_LIST_HEAD(&rb->event_list);
+	spin_lock_init(&rb->event_lock);
+
+	/*
+	 * perf_output_begin() only checks rb->paused, therefore
+	 * rb->paused must be true if we have no pages for output.
+	 */
+	if (!rb->nr_pages)
+		rb->paused = 1;
+}
+
+void perf_aux_output_flag(struct perf_output_handle *handle, u64 flags)
+{
+	/*
+	 * OVERWRITE is determined by perf_aux_output_end() and can't
+	 * be passed in directly.
+	 */
+	if (WARN_ON_ONCE(flags & PERF_AUX_FLAG_OVERWRITE))
+		return;
+
+	handle->aux_flags |= flags;
+}
+EXPORT_SYMBOL_GPL(perf_aux_output_flag);
+
+/*
+ * This is called before hardware starts writing to the AUX area to
+ * obtain an output handle and make sure there's room in the buffer.
+ * When the capture completes, call perf_aux_output_end() to commit
+ * the recorded data to the buffer.
+ *
+ * The ordering is similar to that of perf_output_{begin,end}, with
+ * the exception of (B), which should be taken care of by the pmu
+ * driver, since ordering rules will differ depending on hardware.
+ *
+ * Call this from pmu::start(); see the comment in perf_aux_output_end()
+ * about its use in pmu callbacks. Both can also be called from the PMI
+ * handler if needed.
+ */
+void *perf_aux_output_begin(struct perf_output_handle *handle,
+			    struct perf_event *event)
+{
+	struct perf_event *output_event = event;
+	unsigned long aux_head, aux_tail;
+	struct ring_buffer *rb;
+
+	if (output_event->parent)
+		output_event = output_event->parent;
+
+	/*
+	 * Since this will typically be open across pmu::add/pmu::del, we
+	 * grab ring_buffer's refcount instead of holding rcu read lock
+	 * to make sure it doesn't disappear under us.
+	 */
+	rb = ring_buffer_get(output_event);
+	if (!rb)
+		return NULL;
+
+	if (!rb_has_aux(rb))
+		goto err;
+
+	/*
+	 * If aux_mmap_count is zero, the aux buffer is in perf_mmap_close(),
+	 * about to get freed, so we leave immediately.
+	 *
+	 * Checking rb::aux_mmap_count and rb::refcount has to be done in
+	 * the same order, see perf_mmap_close. Otherwise we end up freeing
+	 * aux pages in this path, which is a bug, because in_atomic().
+	 */
+	if (!atomic_read(&rb->aux_mmap_count))
+		goto err;
+
+	if (!atomic_inc_not_zero(&rb->aux_refcount))
+		goto err;
+
+	/*
+	 * Nesting is not supported for AUX area, make sure nested
+	 * writers are caught early
+	 */
+	if (WARN_ON_ONCE(local_xchg(&rb->aux_nest, 1)))
+		goto err_put;
+
+	aux_head = rb->aux_head;
+
+	handle->rb = rb;
+	handle->event = event;
+	handle->head = aux_head;
+	handle->size = 0;
+	handle->aux_flags = 0;
+
+	/*
+	 * In overwrite mode, AUX data stores do not depend on aux_tail,
+	 * therefore (A) control dependency barrier does not exist. The
+	 * (B) <-> (C) ordering is still observed by the pmu driver.
+	 */
+	if (!rb->aux_overwrite) {
+		aux_tail = READ_ONCE(rb->user_page->aux_tail);
+		handle->wakeup = rb->aux_wakeup + rb->aux_watermark;
+		if (aux_head - aux_tail < perf_aux_size(rb))
+			handle->size = CIRC_SPACE(aux_head, aux_tail, perf_aux_size(rb));
+
+		/*
+		 * handle->size computation depends on aux_tail load; this forms a
+		 * control dependency barrier separating aux_tail load from aux data
+		 * store that will be enabled on successful return
+		 */
+		if (!handle->size) { /* A, matches D */
+			event->pending_disable = smp_processor_id();
+			perf_output_wakeup(handle);
+			local_set(&rb->aux_nest, 0);
+			goto err_put;
+		}
+	}
+
+	return handle->rb->aux_priv;
+
+err_put:
+	/* can't be last */
+	rb_free_aux(rb);
+
+err:
+	ring_buffer_put(rb);
+	handle->event = NULL;
+
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(perf_aux_output_begin);
+
+static __always_inline bool rb_need_aux_wakeup(struct ring_buffer *rb)
+{
+	if (rb->aux_overwrite)
+		return false;
+
+	if (rb->aux_head - rb->aux_wakeup >= rb->aux_watermark) {
+		rb->aux_wakeup = rounddown(rb->aux_head, rb->aux_watermark);
+		return true;
+	}
+
+	return false;
+}
+
+/*
+ * Commit the data written by hardware into the ring buffer by adjusting
+ * aux_head and posting a PERF_RECORD_AUX into the perf buffer. It is the
+ * pmu driver's responsibility to observe ordering rules of the hardware,
+ * so that all the data is externally visible before this is called.
+ *
+ * Note: this has to be called from pmu::stop() callback, as the assumption
+ * of the AUX buffer management code is that after pmu::stop(), the AUX
+ * transaction must be stopped and therefore drop the AUX reference count.
+ */
+void perf_aux_output_end(struct perf_output_handle *handle, unsigned long size)
+{
+	bool wakeup = !!(handle->aux_flags & PERF_AUX_FLAG_TRUNCATED);
+	struct ring_buffer *rb = handle->rb;
+	unsigned long aux_head;
+
+	/* in overwrite mode, driver provides aux_head via handle */
+	if (rb->aux_overwrite) {
+		handle->aux_flags |= PERF_AUX_FLAG_OVERWRITE;
+
+		aux_head = handle->head;
+		rb->aux_head = aux_head;
+	} else {
+		handle->aux_flags &= ~PERF_AUX_FLAG_OVERWRITE;
+
+		aux_head = rb->aux_head;
+		rb->aux_head += size;
+	}
+
+	if (size || handle->aux_flags) {
+		/*
+		 * Only send RECORD_AUX if we have something useful to communicate
+		 */
+
+		perf_event_aux_event(handle->event, aux_head, size,
+		                     handle->aux_flags);
+	}
+
+	WRITE_ONCE(rb->user_page->aux_head, rb->aux_head);
+	if (rb_need_aux_wakeup(rb))
+		wakeup = true;
+
+	if (wakeup) {
+		if (handle->aux_flags & PERF_AUX_FLAG_TRUNCATED)
+			handle->event->pending_disable = smp_processor_id();
+		perf_output_wakeup(handle);
+	}
+
+	handle->event = NULL;
+
+	local_set(&rb->aux_nest, 0);
+	/* can't be last */
+	rb_free_aux(rb);
+	ring_buffer_put(rb);
+}
+EXPORT_SYMBOL_GPL(perf_aux_output_end);
+
+/*
+ * Skip over a given number of bytes in the AUX buffer, due to, for example,
+ * hardware's alignment constraints.
+ */
+int perf_aux_output_skip(struct perf_output_handle *handle, unsigned long size)
+{
+	struct ring_buffer *rb = handle->rb;
+
+	if (size > handle->size)
+		return -ENOSPC;
+
+	rb->aux_head += size;
+
+	WRITE_ONCE(rb->user_page->aux_head, rb->aux_head);
+	if (rb_need_aux_wakeup(rb)) {
+		perf_output_wakeup(handle);
+		handle->wakeup = rb->aux_wakeup + rb->aux_watermark;
+	}
+
+	handle->head = rb->aux_head;
+	handle->size -= size;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(perf_aux_output_skip);
+
+void *perf_get_aux(struct perf_output_handle *handle)
+{
+	/* this is only valid between perf_aux_output_begin and *_end */
+	if (!handle->event)
+		return NULL;
+
+	return handle->rb->aux_priv;
+}
+EXPORT_SYMBOL_GPL(perf_get_aux);
+
+#define PERF_AUX_GFP	(GFP_KERNEL | __GFP_ZERO | __GFP_NOWARN | __GFP_NORETRY)
+
+static struct page *rb_alloc_aux_page(int node, int order)
+{
+	struct page *page;
+
+	if (order > MAX_ORDER)
+		order = MAX_ORDER;
+
+	do {
+		page = alloc_pages_node(node, PERF_AUX_GFP, order);
+	} while (!page && order--);
+
+	if (page && order) {
+		/*
+		 * Communicate the allocation size to the driver:
+		 * if we managed to secure a high-order allocation,
+		 * set its first page's private to this order;
+		 * !PagePrivate(page) means it's just a normal page.
+		 */
+		split_page(page, order);
+		SetPagePrivate(page);
+		set_page_private(page, order);
+	}
+
+	return page;
+}
+
+static void rb_free_aux_page(struct ring_buffer *rb, int idx)
+{
+	struct page *page = virt_to_page(rb->aux_pages[idx]);
+
+	ClearPagePrivate(page);
+	page->mapping = NULL;
+	__free_page(page);
+}
+
+static void __rb_free_aux(struct ring_buffer *rb)
+{
+	int pg;
+
+	/*
+	 * Should never happen, the last reference should be dropped from
+	 * perf_mmap_close() path, which first stops aux transactions (which
+	 * in turn are the atomic holders of aux_refcount) and then does the
+	 * last rb_free_aux().
+	 */
+	WARN_ON_ONCE(in_atomic());
+
+	if (rb->aux_priv) {
+		rb->free_aux(rb->aux_priv);
+		rb->free_aux = NULL;
+		rb->aux_priv = NULL;
+	}
+
+	if (rb->aux_nr_pages) {
+		for (pg = 0; pg < rb->aux_nr_pages; pg++)
+			rb_free_aux_page(rb, pg);
+
+		kfree(rb->aux_pages);
+		rb->aux_nr_pages = 0;
+	}
+}
+
+int rb_alloc_aux(struct ring_buffer *rb, struct perf_event *event,
+		 pgoff_t pgoff, int nr_pages, long watermark, int flags)
+{
+	bool overwrite = !(flags & RING_BUFFER_WRITABLE);
+	int node = (event->cpu == -1) ? -1 : cpu_to_node(event->cpu);
+	int ret = -ENOMEM, max_order = 0;
+
+	if (!has_aux(event))
+		return -EOPNOTSUPP;
+
+	if (event->pmu->capabilities & PERF_PMU_CAP_AUX_NO_SG) {
+		/*
+		 * We need to start with the max_order that fits in nr_pages,
+		 * not the other way around, hence ilog2() and not get_order.
+		 */
+		max_order = ilog2(nr_pages);
+
+		/*
+		 * PMU requests more than one contiguous chunks of memory
+		 * for SW double buffering
+		 */
+		if ((event->pmu->capabilities & PERF_PMU_CAP_AUX_SW_DOUBLEBUF) &&
+		    !overwrite) {
+			if (!max_order)
+				return -EINVAL;
+
+			max_order--;
+		}
+	}
+
+	rb->aux_pages = kcalloc_node(nr_pages, sizeof(void *), GFP_KERNEL,
+				     node);
+	if (!rb->aux_pages)
+		return -ENOMEM;
+
+	rb->free_aux = event->pmu->free_aux;
+	for (rb->aux_nr_pages = 0; rb->aux_nr_pages < nr_pages;) {
+		struct page *page;
+		int last, order;
+
+		order = min(max_order, ilog2(nr_pages - rb->aux_nr_pages));
+		page = rb_alloc_aux_page(node, order);
+		if (!page)
+			goto out;
+
+		for (last = rb->aux_nr_pages + (1 << page_private(page));
+		     last > rb->aux_nr_pages; rb->aux_nr_pages++)
+			rb->aux_pages[rb->aux_nr_pages] = page_address(page++);
+	}
+
+	/*
+	 * In overwrite mode, PMUs that don't support SG may not handle more
+	 * than one contiguous allocation, since they rely on PMI to do double
+	 * buffering. In this case, the entire buffer has to be one contiguous
+	 * chunk.
+	 */
+	if ((event->pmu->capabilities & PERF_PMU_CAP_AUX_NO_SG) &&
+	    overwrite) {
+		struct page *page = virt_to_page(rb->aux_pages[0]);
+
+		if (page_private(page) != max_order)
+			goto out;
+	}
+
+	rb->aux_priv = event->pmu->setup_aux(event, rb->aux_pages, nr_pages,
+					     overwrite);
+	if (!rb->aux_priv)
+		goto out;
+
+	ret = 0;
+
+	/*
+	 * aux_pages (and pmu driver's private data, aux_priv) will be
+	 * referenced in both producer's and consumer's contexts, thus
+	 * we keep a refcount here to make sure either of the two can
+	 * reference them safely.
+	 */
+	atomic_set(&rb->aux_refcount, 1);
+
+	rb->aux_overwrite = overwrite;
+	rb->aux_watermark = watermark;
+
+	if (!rb->aux_watermark && !rb->aux_overwrite)
+		rb->aux_watermark = nr_pages << (PAGE_SHIFT - 1);
+
+out:
+	if (!ret)
+		rb->aux_pgoff = pgoff;
+	else
+		__rb_free_aux(rb);
+
+	return ret;
+}
+
+void rb_free_aux(struct ring_buffer *rb)
+{
+	if (atomic_dec_and_test(&rb->aux_refcount))
+		__rb_free_aux(rb);
+}
+
+#ifndef CONFIG_PERF_USE_VMALLOC
+
+/*
+ * Back perf_mmap() with regular GFP_KERNEL-0 pages.
+ */
+
+static struct page *
+__perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff)
+{
+	if (pgoff > rb->nr_pages)
+		return NULL;
+
+	if (pgoff == 0)
+		return virt_to_page(rb->user_page);
+
+	return virt_to_page(rb->data_pages[pgoff - 1]);
+}
+
+static void *perf_mmap_alloc_page(int cpu)
+{
+	struct page *page;
+	int node;
+
+	node = (cpu == -1) ? cpu : cpu_to_node(cpu);
+	page = alloc_pages_node(node, GFP_KERNEL | __GFP_ZERO, 0);
+	if (!page)
+		return NULL;
+
+	return page_address(page);
+}
+
+struct ring_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags)
+{
+	struct ring_buffer *rb;
+	unsigned long size;
+	int i;
+
+	size = sizeof(struct ring_buffer);
+	size += nr_pages * sizeof(void *);
+
+	if (order_base_2(size) >= PAGE_SHIFT+MAX_ORDER)
+		goto fail;
+
+	rb = kzalloc(size, GFP_KERNEL);
+	if (!rb)
+		goto fail;
+
+	rb->user_page = perf_mmap_alloc_page(cpu);
+	if (!rb->user_page)
+		goto fail_user_page;
+
+	for (i = 0; i < nr_pages; i++) {
+		rb->data_pages[i] = perf_mmap_alloc_page(cpu);
+		if (!rb->data_pages[i])
+			goto fail_data_pages;
+	}
+
+	rb->nr_pages = nr_pages;
+
+	ring_buffer_init(rb, watermark, flags);
+
+	return rb;
+
+fail_data_pages:
+	for (i--; i >= 0; i--)
+		free_page((unsigned long)rb->data_pages[i]);
+
+	free_page((unsigned long)rb->user_page);
+
+fail_user_page:
+	kfree(rb);
+
+fail:
+	return NULL;
+}
+
+static void perf_mmap_free_page(unsigned long addr)
+{
+	struct page *page = virt_to_page((void *)addr);
+
+	page->mapping = NULL;
+	__free_page(page);
+}
+
+void rb_free(struct ring_buffer *rb)
+{
+	int i;
+
+	perf_mmap_free_page((unsigned long)rb->user_page);
+	for (i = 0; i < rb->nr_pages; i++)
+		perf_mmap_free_page((unsigned long)rb->data_pages[i]);
+	kfree(rb);
+}
+
+#else
+static int data_page_nr(struct ring_buffer *rb)
+{
+	return rb->nr_pages << page_order(rb);
+}
+
+static struct page *
+__perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff)
+{
+	/* The '>' counts in the user page. */
+	if (pgoff > data_page_nr(rb))
+		return NULL;
+
+	return vmalloc_to_page((void *)rb->user_page + pgoff * PAGE_SIZE);
+}
+
+static void perf_mmap_unmark_page(void *addr)
+{
+	struct page *page = vmalloc_to_page(addr);
+
+	page->mapping = NULL;
+}
+
+static void rb_free_work(struct work_struct *work)
+{
+	struct ring_buffer *rb;
+	void *base;
+	int i, nr;
+
+	rb = container_of(work, struct ring_buffer, work);
+	nr = data_page_nr(rb);
+
+	base = rb->user_page;
+	/* The '<=' counts in the user page. */
+	for (i = 0; i <= nr; i++)
+		perf_mmap_unmark_page(base + (i * PAGE_SIZE));
+
+	vfree(base);
+	kfree(rb);
+}
+
+void rb_free(struct ring_buffer *rb)
+{
+	schedule_work(&rb->work);
+}
+
+struct ring_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags)
+{
+	struct ring_buffer *rb;
+	unsigned long size;
+	void *all_buf;
+
+	size = sizeof(struct ring_buffer);
+	size += sizeof(void *);
+
+	rb = kzalloc(size, GFP_KERNEL);
+	if (!rb)
+		goto fail;
+
+	INIT_WORK(&rb->work, rb_free_work);
+
+	all_buf = vmalloc_user((nr_pages + 1) * PAGE_SIZE);
+	if (!all_buf)
+		goto fail_all_buf;
+
+	rb->user_page = all_buf;
+	rb->data_pages[0] = all_buf + PAGE_SIZE;
+	if (nr_pages) {
+		rb->nr_pages = 1;
+		rb->page_order = ilog2(nr_pages);
+	}
+
+	ring_buffer_init(rb, watermark, flags);
+
+	return rb;
+
+fail_all_buf:
+	kfree(rb);
+
+fail:
+	return NULL;
+}
+
+#endif
+
+struct page *
+perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff)
+{
+	if (rb->aux_nr_pages) {
+		/* above AUX space */
+		if (pgoff > rb->aux_pgoff + rb->aux_nr_pages)
+			return NULL;
+
+		/* AUX space */
+		if (pgoff >= rb->aux_pgoff) {
+			int aux_pgoff = array_index_nospec(pgoff - rb->aux_pgoff, rb->aux_nr_pages);
+			return virt_to_page(rb->aux_pages[aux_pgoff]);
+		}
+	}
+
+	return __perf_mmap_to_page(rb, pgoff);
+}