path: root/tools/perf/util/thread-stack.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 * thread-stack.c: Synthesize a thread's stack using call / return events
 * Copyright (c) 2014, Intel Corporation.
 */

#include <linux/rbtree.h>
#include <linux/list.h>
#include <linux/log2.h>
#include <linux/zalloc.h>
#include <errno.h>
#include <stdlib.h>
#include <string.h>
#include "thread.h"
#include "event.h"
#include "machine.h"
#include "env.h"
#include "debug.h"
#include "symbol.h"
#include "comm.h"
#include "call-path.h"
#include "thread-stack.h"

#define STACK_GROWTH 2048

/*
 * State of retpoline detection.
 *
 * RETPOLINE_NONE: no retpoline detection
 * X86_RETPOLINE_POSSIBLE: x86 retpoline possible
 * X86_RETPOLINE_DETECTED: x86 retpoline detected
 */
enum retpoline_state_t {
	RETPOLINE_NONE,
	X86_RETPOLINE_POSSIBLE,
	X86_RETPOLINE_DETECTED,
};

/**
 * struct thread_stack_entry - thread stack entry.
 * @ret_addr: return address
 * @timestamp: timestamp (if known)
 * @ref: external reference (e.g. db_id of sample)
 * @branch_count: the branch count when the entry was created
 * @insn_count: the instruction count when the entry was created
 * @cyc_count the cycle count when the entry was created
 * @db_id: id used for db-export
 * @cp: call path
 * @no_call: a 'call' was not seen
 * @trace_end: a 'call' but trace ended
 * @non_call: a branch but not a 'call' to the start of a different symbol
 */
struct thread_stack_entry {
	u64 ret_addr;
	u64 timestamp;
	u64 ref;
	u64 branch_count;
	u64 insn_count;
	u64 cyc_count;
	u64 db_id;
	struct call_path *cp;
	bool no_call;
	bool trace_end;
	bool non_call;
};

/**
 * struct thread_stack - thread stack constructed from 'call' and 'return'
 *                       branch samples.
 * @stack: array that holds the stack
 * @cnt: number of entries in the stack
 * @sz: current maximum stack size
 * @trace_nr: current trace number
 * @branch_count: running branch count
 * @insn_count: running  instruction count
 * @cyc_count running  cycle count
 * @kernel_start: kernel start address
 * @last_time: last timestamp
 * @crp: call/return processor
 * @comm: current comm
 * @arr_sz: size of array if this is the first element of an array
 * @rstate: used to detect retpolines
 * @br_stack_rb: branch stack (ring buffer)
 * @br_stack_sz: maximum branch stack size
 * @br_stack_pos: current position in @br_stack_rb
 * @mispred_all: mark all branches as mispredicted
 */
struct thread_stack {
	struct thread_stack_entry *stack;
	size_t cnt;
	size_t sz;
	u64 trace_nr;
	u64 branch_count;
	u64 insn_count;
	u64 cyc_count;
	u64 kernel_start;
	u64 last_time;
	struct call_return_processor *crp;
	struct comm *comm;
	unsigned int arr_sz;
	enum retpoline_state_t rstate;
	struct branch_stack *br_stack_rb;
	unsigned int br_stack_sz;
	unsigned int br_stack_pos;
	bool mispred_all;
};

/*
 * Assume pid == tid == 0 identifies the idle task as defined by
 * perf_session__register_idle_thread(). The idle task is really 1 task per cpu,
 * and therefore requires a stack for each cpu.
 */
static inline bool thread_stack__per_cpu(struct thread *thread)
{
	return !(thread__tid(thread) || thread__pid(thread));
}

static int thread_stack__grow(struct thread_stack *ts)
{
	struct thread_stack_entry *new_stack;
	size_t sz, new_sz;

	new_sz = ts->sz + STACK_GROWTH;
	sz = new_sz * sizeof(struct thread_stack_entry);

	new_stack = realloc(ts->stack, sz);
	if (!new_stack)
		return -ENOMEM;

	ts->stack = new_stack;
	ts->sz = new_sz;

	return 0;
}

static int thread_stack__init(struct thread_stack *ts, struct thread *thread,
			      struct call_return_processor *crp,
			      bool callstack, unsigned int br_stack_sz)
{
	int err;

	if (callstack) {
		err = thread_stack__grow(ts);
		if (err)
			return err;
	}

	if (br_stack_sz) {
		size_t sz = sizeof(struct branch_stack);

		sz += br_stack_sz * sizeof(struct branch_entry);
		ts->br_stack_rb = zalloc(sz);
		if (!ts->br_stack_rb)
			return -ENOMEM;
		ts->br_stack_sz = br_stack_sz;
	}

	if (thread__maps(t