2019-01-18 01:09:19 +07:00
|
|
|
// SPDX-License-Identifier: GPL-2.0+
|
2016-01-01 09:33:22 +07:00
|
|
|
/*
|
|
|
|
* Read-Copy Update module-based performance-test facility
|
|
|
|
*
|
|
|
|
* Copyright (C) IBM Corporation, 2015
|
|
|
|
*
|
2019-01-18 01:09:19 +07:00
|
|
|
* Authors: Paul E. McKenney <paulmck@linux.ibm.com>
|
2016-01-01 09:33:22 +07:00
|
|
|
*/
|
2018-05-16 02:25:05 +07:00
|
|
|
|
|
|
|
#define pr_fmt(fmt) fmt
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/init.h>
|
2019-12-19 23:22:42 +07:00
|
|
|
#include <linux/mm.h>
|
2016-01-01 09:33:22 +07:00
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/kthread.h>
|
|
|
|
#include <linux/err.h>
|
|
|
|
#include <linux/spinlock.h>
|
|
|
|
#include <linux/smp.h>
|
|
|
|
#include <linux/rcupdate.h>
|
|
|
|
#include <linux/interrupt.h>
|
|
|
|
#include <linux/sched.h>
|
2017-02-02 00:07:51 +07:00
|
|
|
#include <uapi/linux/sched/types.h>
|
2016-01-01 09:33:22 +07:00
|
|
|
#include <linux/atomic.h>
|
|
|
|
#include <linux/bitops.h>
|
|
|
|
#include <linux/completion.h>
|
|
|
|
#include <linux/moduleparam.h>
|
|
|
|
#include <linux/percpu.h>
|
|
|
|
#include <linux/notifier.h>
|
|
|
|
#include <linux/reboot.h>
|
|
|
|
#include <linux/freezer.h>
|
|
|
|
#include <linux/cpu.h>
|
|
|
|
#include <linux/delay.h>
|
|
|
|
#include <linux/stat.h>
|
|
|
|
#include <linux/srcu.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <asm/byteorder.h>
|
|
|
|
#include <linux/torture.h>
|
|
|
|
#include <linux/vmalloc.h>
|
|
|
|
|
2017-05-03 23:51:55 +07:00
|
|
|
#include "rcu.h"
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
MODULE_LICENSE("GPL");
|
2019-01-18 01:09:19 +07:00
|
|
|
MODULE_AUTHOR("Paul E. McKenney <paulmck@linux.ibm.com>");
|
2016-01-01 09:33:22 +07:00
|
|
|
|
|
|
|
#define PERF_FLAG "-perf:"
|
|
|
|
#define PERFOUT_STRING(s) \
|
2016-08-21 14:54:39 +07:00
|
|
|
pr_alert("%s" PERF_FLAG " %s\n", perf_type, s)
|
2016-01-01 09:33:22 +07:00
|
|
|
#define VERBOSE_PERFOUT_STRING(s) \
|
|
|
|
do { if (verbose) pr_alert("%s" PERF_FLAG " %s\n", perf_type, s); } while (0)
|
|
|
|
#define VERBOSE_PERFOUT_ERRSTRING(s) \
|
|
|
|
do { if (verbose) pr_alert("%s" PERF_FLAG "!!! %s\n", perf_type, s); } while (0)
|
|
|
|
|
2018-02-02 10:19:04 +07:00
|
|
|
/*
|
|
|
|
* The intended use cases for the nreaders and nwriters module parameters
|
|
|
|
* are as follows:
|
|
|
|
*
|
|
|
|
* 1. Specify only the nr_cpus kernel boot parameter. This will
|
|
|
|
* set both nreaders and nwriters to the value specified by
|
|
|
|
* nr_cpus for a mixed reader/writer test.
|
|
|
|
*
|
|
|
|
* 2. Specify the nr_cpus kernel boot parameter, but set
|
|
|
|
* rcuperf.nreaders to zero. This will set nwriters to the
|
|
|
|
* value specified by nr_cpus for an update-only test.
|
|
|
|
*
|
|
|
|
* 3. Specify the nr_cpus kernel boot parameter, but set
|
|
|
|
* rcuperf.nwriters to zero. This will set nreaders to the
|
|
|
|
* value specified by nr_cpus for a read-only test.
|
|
|
|
*
|
|
|
|
* Various other use cases may of course be specified.
|
|
|
|
*/
|
|
|
|
|
2018-12-28 22:48:43 +07:00
|
|
|
#ifdef MODULE
|
|
|
|
# define RCUPERF_SHUTDOWN 0
|
|
|
|
#else
|
|
|
|
# define RCUPERF_SHUTDOWN 1
|
|
|
|
#endif
|
|
|
|
|
2017-04-18 02:47:10 +07:00
|
|
|
torture_param(bool, gp_async, false, "Use asynchronous GP wait primitives");
|
|
|
|
torture_param(int, gp_async_max, 1000, "Max # outstanding waits per reader");
|
2016-05-25 08:25:33 +07:00
|
|
|
torture_param(bool, gp_exp, false, "Use expedited GP wait primitives");
|
2016-01-31 11:56:38 +07:00
|
|
|
torture_param(int, holdoff, 10, "Holdoff time before test start (s)");
|
2018-02-02 10:19:04 +07:00
|
|
|
torture_param(int, nreaders, -1, "Number of RCU reader threads");
|
2016-01-01 09:33:22 +07:00
|
|
|
torture_param(int, nwriters, -1, "Number of RCU updater threads");
|
2018-12-28 22:48:43 +07:00
|
|
|
torture_param(bool, shutdown, RCUPERF_SHUTDOWN,
|
2017-04-22 06:09:15 +07:00
|
|
|
"Shutdown at end of performance tests.");
|
2018-05-10 00:29:18 +07:00
|
|
|
torture_param(int, verbose, 1, "Enable verbose debugging printk()s");
|
2017-04-26 05:12:56 +07:00
|
|
|
torture_param(int, writer_holdoff, 0, "Holdoff (us) between GPs, zero to disable");
|
2019-08-30 23:36:29 +07:00
|
|
|
torture_param(int, kfree_rcu_test, 0, "Do we run a kfree_rcu() perf test?");
|
2020-03-16 23:32:26 +07:00
|
|
|
torture_param(int, kfree_mult, 1, "Multiple of kfree_obj size to allocate.");
|
2016-01-01 09:33:22 +07:00
|
|
|
|
|
|
|
static char *perf_type = "rcu";
|
|
|
|
module_param(perf_type, charp, 0444);
|
2019-05-31 22:15:45 +07:00
|
|
|
MODULE_PARM_DESC(perf_type, "Type of RCU to performance-test (rcu, srcu, ...)");
|
2016-01-01 09:33:22 +07:00
|
|
|
|
|
|
|
static int nrealreaders;
|
|
|
|
static int nrealwriters;
|
|
|
|
static struct task_struct **writer_tasks;
|
|
|
|
static struct task_struct **reader_tasks;
|
|
|
|
static struct task_struct *shutdown_task;
|
|
|
|
|
|
|
|
static u64 **writer_durations;
|
|
|
|
static int *writer_n_durations;
|
|
|
|
static atomic_t n_rcu_perf_reader_started;
|
|
|
|
static atomic_t n_rcu_perf_writer_started;
|
|
|
|
static atomic_t n_rcu_perf_writer_finished;
|
|
|
|
static wait_queue_head_t shutdown_wq;
|
|
|
|
static u64 t_rcu_perf_writer_started;
|
|
|
|
static u64 t_rcu_perf_writer_finished;
|
2019-08-30 23:36:29 +07:00
|
|
|
static unsigned long b_rcu_gp_test_started;
|
|
|
|
static unsigned long b_rcu_gp_test_finished;
|
2017-04-18 02:47:10 +07:00
|
|
|
static DEFINE_PER_CPU(atomic_t, n_async_inflight);
|
2016-01-01 09:33:22 +07:00
|
|
|
|
|
|
|
#define MAX_MEAS 10000
|
|
|
|
#define MIN_MEAS 100
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Operations vector for selecting different types of tests.
|
|
|
|
*/
|
|
|
|
|
|
|
|
struct rcu_perf_ops {
|
|
|
|
int ptype;
|
|
|
|
void (*init)(void);
|
|
|
|
void (*cleanup)(void);
|
|
|
|
int (*readlock)(void);
|
|
|
|
void (*readunlock)(int idx);
|
2018-04-28 01:39:34 +07:00
|
|
|
unsigned long (*get_gp_seq)(void);
|
2018-05-16 05:24:41 +07:00
|
|
|
unsigned long (*gp_diff)(unsigned long new, unsigned long old);
|
2016-01-01 09:33:22 +07:00
|
|
|
unsigned long (*exp_completed)(void);
|
2017-04-18 02:47:10 +07:00
|
|
|
void (*async)(struct rcu_head *head, rcu_callback_t func);
|
|
|
|
void (*gp_barrier)(void);
|
2016-01-01 09:33:22 +07:00
|
|
|
void (*sync)(void);
|
|
|
|
void (*exp_sync)(void);
|
|
|
|
const char *name;
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct rcu_perf_ops *cur_ops;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Definitions for rcu perf testing.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int rcu_perf_read_lock(void) __acquires(RCU)
|
|
|
|
{
|
|
|
|
rcu_read_lock();
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void rcu_perf_read_unlock(int idx) __releases(RCU)
|
|
|
|
{
|
|
|
|
rcu_read_unlock();
|
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned long __maybe_unused rcu_no_completed(void)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void rcu_sync_perf_init(void)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct rcu_perf_ops rcu_ops = {
|
|
|
|
.ptype = RCU_FLAVOR,
|
|
|
|
.init = rcu_sync_perf_init,
|
|
|
|
.readlock = rcu_perf_read_lock,
|
|
|
|
.readunlock = rcu_perf_read_unlock,
|
2018-04-28 01:39:34 +07:00
|
|
|
.get_gp_seq = rcu_get_gp_seq,
|
2018-05-16 05:24:41 +07:00
|
|
|
.gp_diff = rcu_seq_diff,
|
2016-01-01 09:33:22 +07:00
|
|
|
.exp_completed = rcu_exp_batches_completed,
|
2017-04-18 02:47:10 +07:00
|
|
|
.async = call_rcu,
|
|
|
|
.gp_barrier = rcu_barrier,
|
2016-01-01 09:33:22 +07:00
|
|
|
.sync = synchronize_rcu,
|
|
|
|
.exp_sync = synchronize_rcu_expedited,
|
|
|
|
.name = "rcu"
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Definitions for srcu perf testing.
|
|
|
|
*/
|
|
|
|
|
|
|
|
DEFINE_STATIC_SRCU(srcu_ctl_perf);
|
|
|
|
static struct srcu_struct *srcu_ctlp = &srcu_ctl_perf;
|
|
|
|
|
|
|
|
static int srcu_perf_read_lock(void) __acquires(srcu_ctlp)
|
|
|
|
{
|
|
|
|
return srcu_read_lock(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void srcu_perf_read_unlock(int idx) __releases(srcu_ctlp)
|
|
|
|
{
|
|
|
|
srcu_read_unlock(srcu_ctlp, idx);
|
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned long srcu_perf_completed(void)
|
|
|
|
{
|
|
|
|
return srcu_batches_completed(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
2017-04-18 02:47:10 +07:00
|
|
|
static void srcu_call_rcu(struct rcu_head *head, rcu_callback_t func)
|
|
|
|
{
|
|
|
|
call_srcu(srcu_ctlp, head, func);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void srcu_rcu_barrier(void)
|
|
|
|
{
|
|
|
|
srcu_barrier(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
static void srcu_perf_synchronize(void)
|
|
|
|
{
|
|
|
|
synchronize_srcu(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void srcu_perf_synchronize_expedited(void)
|
|
|
|
{
|
|
|
|
synchronize_srcu_expedited(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct rcu_perf_ops srcu_ops = {
|
|
|
|
.ptype = SRCU_FLAVOR,
|
|
|
|
.init = rcu_sync_perf_init,
|
|
|
|
.readlock = srcu_perf_read_lock,
|
|
|
|
.readunlock = srcu_perf_read_unlock,
|
2018-04-28 01:39:34 +07:00
|
|
|
.get_gp_seq = srcu_perf_completed,
|
2018-05-16 05:24:41 +07:00
|
|
|
.gp_diff = rcu_seq_diff,
|
2016-01-01 09:33:22 +07:00
|
|
|
.exp_completed = srcu_perf_completed,
|
2017-04-18 02:47:10 +07:00
|
|
|
.async = srcu_call_rcu,
|
|
|
|
.gp_barrier = srcu_rcu_barrier,
|
2016-01-01 09:33:22 +07:00
|
|
|
.sync = srcu_perf_synchronize,
|
|
|
|
.exp_sync = srcu_perf_synchronize_expedited,
|
|
|
|
.name = "srcu"
|
|
|
|
};
|
|
|
|
|
2017-04-20 03:43:21 +07:00
|
|
|
static struct srcu_struct srcud;
|
|
|
|
|
|
|
|
static void srcu_sync_perf_init(void)
|
|
|
|
{
|
|
|
|
srcu_ctlp = &srcud;
|
|
|
|
init_srcu_struct(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void srcu_sync_perf_cleanup(void)
|
|
|
|
{
|
|
|
|
cleanup_srcu_struct(srcu_ctlp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct rcu_perf_ops srcud_ops = {
|
|
|
|
.ptype = SRCU_FLAVOR,
|
|
|
|
.init = srcu_sync_perf_init,
|
|
|
|
.cleanup = srcu_sync_perf_cleanup,
|
|
|
|
.readlock = srcu_perf_read_lock,
|
|
|
|
.readunlock = srcu_perf_read_unlock,
|
2018-04-28 01:39:34 +07:00
|
|
|
.get_gp_seq = srcu_perf_completed,
|
2018-05-16 05:24:41 +07:00
|
|
|
.gp_diff = rcu_seq_diff,
|
2017-04-20 03:43:21 +07:00
|
|
|
.exp_completed = srcu_perf_completed,
|
|
|
|
.async = srcu_call_rcu,
|
|
|
|
.gp_barrier = srcu_rcu_barrier,
|
|
|
|
.sync = srcu_perf_synchronize,
|
|
|
|
.exp_sync = srcu_perf_synchronize_expedited,
|
|
|
|
.name = "srcud"
|
|
|
|
};
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
/*
|
|
|
|
* Definitions for RCU-tasks perf testing.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int tasks_perf_read_lock(void)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void tasks_perf_read_unlock(int idx)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct rcu_perf_ops tasks_ops = {
|
|
|
|
.ptype = RCU_TASKS_FLAVOR,
|
|
|
|
.init = rcu_sync_perf_init,
|
|
|
|
.readlock = tasks_perf_read_lock,
|
|
|
|
.readunlock = tasks_perf_read_unlock,
|
2018-04-28 01:39:34 +07:00
|
|
|
.get_gp_seq = rcu_no_completed,
|
2018-05-16 05:24:41 +07:00
|
|
|
.gp_diff = rcu_seq_diff,
|
2017-04-18 02:47:10 +07:00
|
|
|
.async = call_rcu_tasks,
|
|
|
|
.gp_barrier = rcu_barrier_tasks,
|
2016-01-01 09:33:22 +07:00
|
|
|
.sync = synchronize_rcu_tasks,
|
|
|
|
.exp_sync = synchronize_rcu_tasks,
|
|
|
|
.name = "tasks"
|
|
|
|
};
|
|
|
|
|
2018-05-16 05:24:41 +07:00
|
|
|
static unsigned long rcuperf_seq_diff(unsigned long new, unsigned long old)
|
|
|
|
{
|
|
|
|
if (!cur_ops->gp_diff)
|
|
|
|
return new - old;
|
|
|
|
return cur_ops->gp_diff(new, old);
|
|
|
|
}
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
/*
|
|
|
|
* If performance tests complete, wait for shutdown to commence.
|
|
|
|
*/
|
|
|
|
static void rcu_perf_wait_shutdown(void)
|
|
|
|
{
|
2018-03-03 07:35:27 +07:00
|
|
|
cond_resched_tasks_rcu_qs();
|
2016-01-01 09:33:22 +07:00
|
|
|
if (atomic_read(&n_rcu_perf_writer_finished) < nrealwriters)
|
|
|
|
return;
|
|
|
|
while (!torture_must_stop())
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* RCU perf reader kthread. Repeatedly does empty RCU read-side
|
|
|
|
* critical section, minimizing update-side interference.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
rcu_perf_reader(void *arg)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
int idx;
|
2016-01-13 05:15:40 +07:00
|
|
|
long me = (long)arg;
|
2016-01-01 09:33:22 +07:00
|
|
|
|
|
|
|
VERBOSE_PERFOUT_STRING("rcu_perf_reader task started");
|
2016-01-13 05:15:40 +07:00
|
|
|
set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids));
|
2016-01-01 09:33:22 +07:00
|
|
|
set_user_nice(current, MAX_NICE);
|
|
|
|
atomic_inc(&n_rcu_perf_reader_started);
|
|
|
|
|
|
|
|
do {
|
|
|
|
local_irq_save(flags);
|
|
|
|
idx = cur_ops->readlock();
|
|
|
|
cur_ops->readunlock(idx);
|
|
|
|
local_irq_restore(flags);
|
|
|
|
rcu_perf_wait_shutdown();
|
|
|
|
} while (!torture_must_stop());
|
|
|
|
torture_kthread_stopping("rcu_perf_reader");
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2017-04-18 02:47:10 +07:00
|
|
|
/*
|
|
|
|
* Callback function for asynchronous grace periods from rcu_perf_writer().
|
|
|
|
*/
|
|
|
|
static void rcu_perf_async_cb(struct rcu_head *rhp)
|
|
|
|
{
|
|
|
|
atomic_dec(this_cpu_ptr(&n_async_inflight));
|
|
|
|
kfree(rhp);
|
|
|
|
}
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
/*
|
|
|
|
* RCU perf writer kthread. Repeatedly does a grace period.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
rcu_perf_writer(void *arg)
|
|
|
|
{
|
|
|
|
int i = 0;
|
|
|
|
int i_max;
|
|
|
|
long me = (long)arg;
|
2017-04-18 02:47:10 +07:00
|
|
|
struct rcu_head *rhp = NULL;
|
2016-01-13 06:17:21 +07:00
|
|
|
struct sched_param sp;
|
2016-01-01 09:33:22 +07:00
|
|
|
bool started = false, done = false, alldone = false;
|
|
|
|
u64 t;
|
|
|
|
u64 *wdp;
|
|
|
|
u64 *wdpp = writer_durations[me];
|
|
|
|
|
|
|
|
VERBOSE_PERFOUT_STRING("rcu_perf_writer task started");
|
|
|
|
WARN_ON(!wdpp);
|
2016-01-13 05:15:40 +07:00
|
|
|
set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids));
|
2016-01-13 06:17:21 +07:00
|
|
|
sp.sched_priority = 1;
|
|
|
|
sched_setscheduler_nocheck(current, SCHED_FIFO, &sp);
|
2016-01-31 11:56:38 +07:00
|
|
|
|
|
|
|
if (holdoff)
|
|
|
|
schedule_timeout_uninterruptible(holdoff * HZ);
|
|
|
|
|
2019-07-04 11:34:30 +07:00
|
|
|
/*
|
|
|
|
* Wait until rcu_end_inkernel_boot() is called for normal GP tests
|
|
|
|
* so that RCU is not always expedited for normal GP tests.
|
|
|
|
* The system_state test is approximate, but works well in practice.
|
|
|
|
*/
|
|
|
|
while (!gp_exp && system_state != SYSTEM_RUNNING)
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
t = ktime_get_mono_fast_ns();
|
|
|
|
if (atomic_inc_return(&n_rcu_perf_writer_started) >= nrealwriters) {
|
|
|
|
t_rcu_perf_writer_started = t;
|
|
|
|
if (gp_exp) {
|
2019-08-30 23:36:29 +07:00
|
|
|
b_rcu_gp_test_started =
|
2016-01-01 09:33:22 +07:00
|
|
|
cur_ops->exp_completed() / 2;
|
|
|
|
} else {
|
2019-08-30 23:36:29 +07:00
|
|
|
b_rcu_gp_test_started = cur_ops->get_gp_seq();
|
2016-01-01 09:33:22 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
do {
|
2017-04-26 05:12:56 +07:00
|
|
|
if (writer_holdoff)
|
|
|
|
udelay(writer_holdoff);
|
2016-01-01 09:33:22 +07:00
|
|
|
wdp = &wdpp[i];
|
|
|
|
*wdp = ktime_get_mono_fast_ns();
|
2017-04-18 02:47:10 +07:00
|
|
|
if (gp_async) {
|
|
|
|
retry:
|
|
|
|
if (!rhp)
|
|
|
|
rhp = kmalloc(sizeof(*rhp), GFP_KERNEL);
|
|
|
|
if (rhp && atomic_read(this_cpu_ptr(&n_async_inflight)) < gp_async_max) {
|
|
|
|
atomic_inc(this_cpu_ptr(&n_async_inflight));
|
|
|
|
cur_ops->async(rhp, rcu_perf_async_cb);
|
|
|
|
rhp = NULL;
|
|
|
|
} else if (!kthread_should_stop()) {
|
|
|
|
cur_ops->gp_barrier();
|
|
|
|
goto retry;
|
|
|
|
} else {
|
|
|
|
kfree(rhp); /* Because we are stopping. */
|
|
|
|
}
|
|
|
|
} else if (gp_exp) {
|
2016-01-01 09:33:22 +07:00
|
|
|
cur_ops->exp_sync();
|
|
|
|
} else {
|
|
|
|
cur_ops->sync();
|
|
|
|
}
|
|
|
|
t = ktime_get_mono_fast_ns();
|
|
|
|
*wdp = t - *wdp;
|
|
|
|
i_max = i;
|
|
|
|
if (!started &&
|
|
|
|
atomic_read(&n_rcu_perf_writer_started) >= nrealwriters)
|
|
|
|
started = true;
|
|
|
|
if (!done && i >= MIN_MEAS) {
|
|
|
|
done = true;
|
2016-01-31 12:32:09 +07:00
|
|
|
sp.sched_priority = 0;
|
|
|
|
sched_setscheduler_nocheck(current,
|
|
|
|
SCHED_NORMAL, &sp);
|
2016-08-21 14:54:39 +07:00
|
|
|
pr_alert("%s%s rcu_perf_writer %ld has %d measurements\n",
|
|
|
|
perf_type, PERF_FLAG, me, MIN_MEAS);
|
2016-01-01 09:33:22 +07:00
|
|
|
if (atomic_inc_return(&n_rcu_perf_writer_finished) >=
|
|
|
|
nrealwriters) {
|
2016-01-31 12:32:09 +07:00
|
|
|
schedule_timeout_interruptible(10);
|
2016-01-30 05:58:17 +07:00
|
|
|
rcu_ftrace_dump(DUMP_ALL);
|
2016-01-01 09:33:22 +07:00
|
|
|
PERFOUT_STRING("Test complete");
|
|
|
|
t_rcu_perf_writer_finished = t;
|
|
|
|
if (gp_exp) {
|
2019-08-30 23:36:29 +07:00
|
|
|
b_rcu_gp_test_finished =
|
2016-01-01 09:33:22 +07:00
|
|
|
cur_ops->exp_completed() / 2;
|
|
|
|
} else {
|
2019-08-30 23:36:29 +07:00
|
|
|
b_rcu_gp_test_finished =
|
2018-04-28 01:39:34 +07:00
|
|
|
cur_ops->get_gp_seq();
|
2016-01-01 09:33:22 +07:00
|
|
|
}
|
2016-02-07 19:31:39 +07:00
|
|
|
if (shutdown) {
|
|
|
|
smp_mb(); /* Assign before wake. */
|
|
|
|
wake_up(&shutdown_wq);
|
|
|
|
}
|
2016-01-01 09:33:22 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
if (done && !alldone &&
|
|
|
|
atomic_read(&n_rcu_perf_writer_finished) >= nrealwriters)
|
|
|
|
alldone = true;
|
|
|
|
if (started && !alldone && i < MAX_MEAS - 1)
|
|
|
|
i++;
|
|
|
|
rcu_perf_wait_shutdown();
|
|
|
|
} while (!torture_must_stop());
|
2017-04-18 02:47:10 +07:00
|
|
|
if (gp_async) {
|
|
|
|
cur_ops->gp_barrier();
|
|
|
|
}
|
2016-01-01 09:33:22 +07:00
|
|
|
writer_n_durations[me] = i_max;
|
|
|
|
torture_kthread_stopping("rcu_perf_writer");
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2018-05-18 01:33:17 +07:00
|
|
|
static void
|
2016-01-01 09:33:22 +07:00
|
|
|
rcu_perf_print_module_parms(struct rcu_perf_ops *cur_ops, const char *tag)
|
|
|
|
{
|
|
|
|
pr_alert("%s" PERF_FLAG
|
|
|
|
"--- %s: nreaders=%d nwriters=%d verbose=%d shutdown=%d\n",
|
|
|
|
perf_type, tag, nrealreaders, nrealwriters, verbose, shutdown);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
rcu_perf_cleanup(void)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
int j;
|
|
|
|
int ngps = 0;
|
|
|
|
u64 *wdp;
|
|
|
|
u64 *wdpp;
|
|
|
|
|
2017-04-15 06:12:52 +07:00
|
|
|
/*
|
|
|
|
* Would like warning at start, but everything is expedited
|
|
|
|
* during the mid-boot phase, so have to wait till the end.
|
|
|
|
*/
|
|
|
|
if (rcu_gp_is_expedited() && !rcu_gp_is_normal() && !gp_exp)
|
|
|
|
VERBOSE_PERFOUT_ERRSTRING("All grace periods expedited, no normal ones to measure!");
|
|
|
|
if (rcu_gp_is_normal() && gp_exp)
|
|
|
|
VERBOSE_PERFOUT_ERRSTRING("All grace periods normal, no expedited ones to measure!");
|
2017-04-18 02:47:10 +07:00
|
|
|
if (gp_exp && gp_async)
|
|
|
|
VERBOSE_PERFOUT_ERRSTRING("No expedited async GPs, so went with async!");
|
2017-04-15 06:12:52 +07:00
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
if (torture_cleanup_begin())
|
|
|
|
return;
|
2019-03-22 00:26:41 +07:00
|
|
|
if (!cur_ops) {
|
|
|
|
torture_cleanup_end();
|
|
|
|
return;
|
|
|
|
}
|
2016-01-01 09:33:22 +07:00
|
|
|
|
|
|
|
if (reader_tasks) {
|
|
|
|
for (i = 0; i < nrealreaders; i++)
|
|
|
|
torture_stop_kthread(rcu_perf_reader,
|
|
|
|
reader_tasks[i]);
|
|
|
|
kfree(reader_tasks);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (writer_tasks) {
|
|
|
|
for (i = 0; i < nrealwriters; i++) {
|
|
|
|
torture_stop_kthread(rcu_perf_writer,
|
|
|
|
writer_tasks[i]);
|
|
|
|
if (!writer_n_durations)
|
|
|
|
continue;
|
|
|
|
j = writer_n_durations[i];
|
|
|
|
pr_alert("%s%s writer %d gps: %d\n",
|
|
|
|
perf_type, PERF_FLAG, i, j);
|
|
|
|
ngps += j;
|
|
|
|
}
|
|
|
|
pr_alert("%s%s start: %llu end: %llu duration: %llu gps: %d batches: %ld\n",
|
|
|
|
perf_type, PERF_FLAG,
|
|
|
|
t_rcu_perf_writer_started, t_rcu_perf_writer_finished,
|
|
|
|
t_rcu_perf_writer_finished -
|
|
|
|
t_rcu_perf_writer_started,
|
|
|
|
ngps,
|
2019-08-30 23:36:29 +07:00
|
|
|
rcuperf_seq_diff(b_rcu_gp_test_finished,
|
|
|
|
b_rcu_gp_test_started));
|
2016-01-01 09:33:22 +07:00
|
|
|
for (i = 0; i < nrealwriters; i++) {
|
|
|
|
if (!writer_durations)
|
|
|
|
break;
|
|
|
|
if (!writer_n_durations)
|
|
|
|
continue;
|
|
|
|
wdpp = writer_durations[i];
|
|
|
|
if (!wdpp)
|
|
|
|
continue;
|
|
|
|
for (j = 0; j <= writer_n_durations[i]; j++) {
|
|
|
|
wdp = &wdpp[j];
|
|
|
|
pr_alert("%s%s %4d writer-duration: %5d %llu\n",
|
|
|
|
perf_type, PERF_FLAG,
|
|
|
|
i, j, *wdp);
|
|
|
|
if (j % 100 == 0)
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
}
|
|
|
|
kfree(writer_durations[i]);
|
|
|
|
}
|
|
|
|
kfree(writer_tasks);
|
|
|
|
kfree(writer_durations);
|
|
|
|
kfree(writer_n_durations);
|
|
|
|
}
|
|
|
|
|
2018-07-08 08:25:10 +07:00
|
|
|
/* Do torture-type-specific cleanup operations. */
|
2016-01-01 09:33:22 +07:00
|
|
|
if (cur_ops->cleanup != NULL)
|
|
|
|
cur_ops->cleanup();
|
|
|
|
|
|
|
|
torture_cleanup_end();
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return the number if non-negative. If -1, the number of CPUs.
|
|
|
|
* If less than -1, that much less than the number of CPUs, but
|
|
|
|
* at least one.
|
|
|
|
*/
|
|
|
|
static int compute_real(int n)
|
|
|
|
{
|
|
|
|
int nr;
|
|
|
|
|
|
|
|
if (n >= 0) {
|
|
|
|
nr = n;
|
|
|
|
} else {
|
|
|
|
nr = num_online_cpus() + 1 + n;
|
|
|
|
if (nr <= 0)
|
|
|
|
nr = 1;
|
|
|
|
}
|
|
|
|
return nr;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* RCU perf shutdown kthread. Just waits to be awakened, then shuts
|
|
|
|
* down system.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
rcu_perf_shutdown(void *arg)
|
|
|
|
{
|
|
|
|
do {
|
|
|
|
wait_event(shutdown_wq,
|
|
|
|
atomic_read(&n_rcu_perf_writer_finished) >=
|
|
|
|
nrealwriters);
|
|
|
|
} while (atomic_read(&n_rcu_perf_writer_finished) < nrealwriters);
|
|
|
|
smp_mb(); /* Wake before output. */
|
|
|
|
rcu_perf_cleanup();
|
|
|
|
kernel_power_off();
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2019-08-30 23:36:29 +07:00
|
|
|
/*
|
|
|
|
* kfree_rcu() performance tests: Start a kfree_rcu() loop on all CPUs for number
|
|
|
|
* of iterations and measure total time and number of GP for all iterations to complete.
|
|
|
|
*/
|
|
|
|
|
|
|
|
torture_param(int, kfree_nthreads, -1, "Number of threads running loops of kfree_rcu().");
|
|
|
|
torture_param(int, kfree_alloc_num, 8000, "Number of allocations and frees done in an iteration.");
|
|
|
|
torture_param(int, kfree_loops, 10, "Number of loops doing kfree_alloc_num allocations and frees.");
|
|
|
|
|
|
|
|
static struct task_struct **kfree_reader_tasks;
|
|
|
|
static int kfree_nrealthreads;
|
|
|
|
static atomic_t n_kfree_perf_thread_started;
|
|
|
|
static atomic_t n_kfree_perf_thread_ended;
|
|
|
|
|
|
|
|
struct kfree_obj {
|
|
|
|
char kfree_obj[8];
|
|
|
|
struct rcu_head rh;
|
|
|
|
};
|
|
|
|
|
|
|
|
static int
|
|
|
|
kfree_perf_thread(void *arg)
|
|
|
|
{
|
|
|
|
int i, loop = 0;
|
|
|
|
long me = (long)arg;
|
|
|
|
struct kfree_obj *alloc_ptr;
|
|
|
|
u64 start_time, end_time;
|
2019-12-19 23:22:42 +07:00
|
|
|
long long mem_begin, mem_during = 0;
|
2019-08-30 23:36:29 +07:00
|
|
|
|
|
|
|
VERBOSE_PERFOUT_STRING("kfree_perf_thread task started");
|
|
|
|
set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids));
|
|
|
|
set_user_nice(current, MAX_NICE);
|
|
|
|
|
|
|
|
start_time = ktime_get_mono_fast_ns();
|
|
|
|
|
|
|
|
if (atomic_inc_return(&n_kfree_perf_thread_started) >= kfree_nrealthreads) {
|
|
|
|
if (gp_exp)
|
|
|
|
b_rcu_gp_test_started = cur_ops->exp_completed() / 2;
|
|
|
|
else
|
|
|
|
b_rcu_gp_test_started = cur_ops->get_gp_seq();
|
|
|
|
}
|
|
|
|
|
|
|
|
do {
|
2019-12-19 23:22:42 +07:00
|
|
|
if (!mem_during) {
|
|
|
|
mem_during = mem_begin = si_mem_available();
|
|
|
|
} else if (loop % (kfree_loops / 4) == 0) {
|
|
|
|
mem_during = (mem_during + si_mem_available()) / 2;
|
|
|
|
}
|
|
|
|
|
2019-08-30 23:36:29 +07:00
|
|
|
for (i = 0; i < kfree_alloc_num; i++) {
|
2020-03-16 23:32:26 +07:00
|
|
|
alloc_ptr = kmalloc(kfree_mult * sizeof(struct kfree_obj), GFP_KERNEL);
|
2019-08-30 23:36:29 +07:00
|
|
|
if (!alloc_ptr)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2019-08-30 23:36:33 +07:00
|
|
|
kfree_rcu(alloc_ptr, rh);
|
2019-08-30 23:36:29 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
cond_resched();
|
|
|
|
} while (!torture_must_stop() && ++loop < kfree_loops);
|
|
|
|
|
|
|
|
if (atomic_inc_return(&n_kfree_perf_thread_ended) >= kfree_nrealthreads) {
|
|
|
|
end_time = ktime_get_mono_fast_ns();
|
|
|
|
|
|
|
|
if (gp_exp)
|
|
|
|
b_rcu_gp_test_finished = cur_ops->exp_completed() / 2;
|
|
|
|
else
|
|
|
|
b_rcu_gp_test_finished = cur_ops->get_gp_seq();
|
|
|
|
|
2019-12-19 23:22:42 +07:00
|
|
|
pr_alert("Total time taken by all kfree'ers: %llu ns, loops: %d, batches: %ld, memory footprint: %lldMB\n",
|
2019-08-30 23:36:29 +07:00
|
|
|
(unsigned long long)(end_time - start_time), kfree_loops,
|
2019-12-19 23:22:42 +07:00
|
|
|
rcuperf_seq_diff(b_rcu_gp_test_finished, b_rcu_gp_test_started),
|
|
|
|
(mem_begin - mem_during) >> (20 - PAGE_SHIFT));
|
|
|
|
|
2019-08-30 23:36:29 +07:00
|
|
|
if (shutdown) {
|
|
|
|
smp_mb(); /* Assign before wake. */
|
|
|
|
wake_up(&shutdown_wq);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
torture_kthread_stopping("kfree_perf_thread");
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
kfree_perf_cleanup(void)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (torture_cleanup_begin())
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (kfree_reader_tasks) {
|
|
|
|
for (i = 0; i < kfree_nrealthreads; i++)
|
|
|
|
torture_stop_kthread(kfree_perf_thread,
|
|
|
|
kfree_reader_tasks[i]);
|
|
|
|
kfree(kfree_reader_tasks);
|
|
|
|
}
|
|
|
|
|
|
|
|
torture_cleanup_end();
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* shutdown kthread. Just waits to be awakened, then shuts down system.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
kfree_perf_shutdown(void *arg)
|
|
|
|
{
|
|
|
|
do {
|
|
|
|
wait_event(shutdown_wq,
|
|
|
|
atomic_read(&n_kfree_perf_thread_ended) >=
|
|
|
|
kfree_nrealthreads);
|
|
|
|
} while (atomic_read(&n_kfree_perf_thread_ended) < kfree_nrealthreads);
|
|
|
|
|
|
|
|
smp_mb(); /* Wake before output. */
|
|
|
|
|
|
|
|
kfree_perf_cleanup();
|
|
|
|
kernel_power_off();
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __init
|
|
|
|
kfree_perf_init(void)
|
|
|
|
{
|
|
|
|
long i;
|
|
|
|
int firsterr = 0;
|
|
|
|
|
|
|
|
kfree_nrealthreads = compute_real(kfree_nthreads);
|
|
|
|
/* Start up the kthreads. */
|
|
|
|
if (shutdown) {
|
|
|
|
init_waitqueue_head(&shutdown_wq);
|
|
|
|
firsterr = torture_create_kthread(kfree_perf_shutdown, NULL,
|
|
|
|
shutdown_task);
|
|
|
|
if (firsterr)
|
|
|
|
goto unwind;
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
}
|
|
|
|
|
2020-03-16 23:32:26 +07:00
|
|
|
pr_alert("kfree object size=%lu\n", kfree_mult * sizeof(struct kfree_obj));
|
|
|
|
|
2019-08-30 23:36:29 +07:00
|
|
|
kfree_reader_tasks = kcalloc(kfree_nrealthreads, sizeof(kfree_reader_tasks[0]),
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (kfree_reader_tasks == NULL) {
|
|
|
|
firsterr = -ENOMEM;
|
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < kfree_nrealthreads; i++) {
|
|
|
|
firsterr = torture_create_kthread(kfree_perf_thread, (void *)i,
|
|
|
|
kfree_reader_tasks[i]);
|
|
|
|
if (firsterr)
|
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
|
|
|
|
while (atomic_read(&n_kfree_perf_thread_started) < kfree_nrealthreads)
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
|
|
|
|
torture_init_end();
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
unwind:
|
|
|
|
torture_init_end();
|
|
|
|
kfree_perf_cleanup();
|
|
|
|
return firsterr;
|
|
|
|
}
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
static int __init
|
|
|
|
rcu_perf_init(void)
|
|
|
|
{
|
|
|
|
long i;
|
|
|
|
int firsterr = 0;
|
|
|
|
static struct rcu_perf_ops *perf_ops[] = {
|
2018-07-08 08:25:10 +07:00
|
|
|
&rcu_ops, &srcu_ops, &srcud_ops, &tasks_ops,
|
2016-01-01 09:33:22 +07:00
|
|
|
};
|
|
|
|
|
2017-11-22 11:19:17 +07:00
|
|
|
if (!torture_init_begin(perf_type, verbose))
|
2016-01-01 09:33:22 +07:00
|
|
|
return -EBUSY;
|
|
|
|
|
|
|
|
/* Process args and tell the world that the perf'er is on the job. */
|
|
|
|
for (i = 0; i < ARRAY_SIZE(perf_ops); i++) {
|
|
|
|
cur_ops = perf_ops[i];
|
|
|
|
if (strcmp(perf_type, cur_ops->name) == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (i == ARRAY_SIZE(perf_ops)) {
|
2018-05-15 03:27:33 +07:00
|
|
|
pr_alert("rcu-perf: invalid perf type: \"%s\"\n", perf_type);
|
2016-01-01 09:33:22 +07:00
|
|
|
pr_alert("rcu-perf types:");
|
|
|
|
for (i = 0; i < ARRAY_SIZE(perf_ops); i++)
|
2018-05-15 03:27:33 +07:00
|
|
|
pr_cont(" %s", perf_ops[i]->name);
|
|
|
|
pr_cont("\n");
|
2018-07-08 08:26:50 +07:00
|
|
|
WARN_ON(!IS_MODULE(CONFIG_RCU_PERF_TEST));
|
2016-01-01 09:33:22 +07:00
|
|
|
firsterr = -EINVAL;
|
2019-03-22 00:26:41 +07:00
|
|
|
cur_ops = NULL;
|
2016-01-01 09:33:22 +07:00
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
if (cur_ops->init)
|
|
|
|
cur_ops->init();
|
|
|
|
|
2019-08-30 23:36:29 +07:00
|
|
|
if (kfree_rcu_test)
|
|
|
|
return kfree_perf_init();
|
|
|
|
|
2016-01-01 09:33:22 +07:00
|
|
|
nrealwriters = compute_real(nwriters);
|
|
|
|
nrealreaders = compute_real(nreaders);
|
|
|
|
atomic_set(&n_rcu_perf_reader_started, 0);
|
|
|
|
atomic_set(&n_rcu_perf_writer_started, 0);
|
|
|
|
atomic_set(&n_rcu_perf_writer_finished, 0);
|
|
|
|
rcu_perf_print_module_parms(cur_ops, "Start of test");
|
|
|
|
|
|
|
|
/* Start up the kthreads. */
|
|
|
|
|
|
|
|
if (shutdown) {
|
|
|
|
init_waitqueue_head(&shutdown_wq);
|
|
|
|
firsterr = torture_create_kthread(rcu_perf_shutdown, NULL,
|
|
|
|
shutdown_task);
|
|
|
|
if (firsterr)
|
|
|
|
goto unwind;
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
}
|
|
|
|
reader_tasks = kcalloc(nrealreaders, sizeof(reader_tasks[0]),
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (reader_tasks == NULL) {
|
|
|
|
VERBOSE_PERFOUT_ERRSTRING("out of memory");
|
|
|
|
firsterr = -ENOMEM;
|
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
for (i = 0; i < nrealreaders; i++) {
|
2016-01-13 05:15:40 +07:00
|
|
|
firsterr = torture_create_kthread(rcu_perf_reader, (void *)i,
|
2016-01-01 09:33:22 +07:00
|
|
|
reader_tasks[i]);
|
|
|
|
if (firsterr)
|
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
while (atomic_read(&n_rcu_perf_reader_started) < nrealreaders)
|
|
|
|
schedule_timeout_uninterruptible(1);
|
|
|
|
writer_tasks = kcalloc(nrealwriters, sizeof(reader_tasks[0]),
|
|
|
|
GFP_KERNEL);
|
|
|
|
writer_durations = kcalloc(nrealwriters, sizeof(*writer_durations),
|
|
|
|
GFP_KERNEL);
|
|
|
|
writer_n_durations =
|
|
|
|
kcalloc(nrealwriters, sizeof(*writer_n_durations),
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (!writer_tasks || !writer_durations || !writer_n_durations) {
|
|
|
|
VERBOSE_PERFOUT_ERRSTRING("out of memory");
|
|
|
|
firsterr = -ENOMEM;
|
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
for (i = 0; i < nrealwriters; i++) {
|
|
|
|
writer_durations[i] =
|
|
|
|
kcalloc(MAX_MEAS, sizeof(*writer_durations[i]),
|
|
|
|
GFP_KERNEL);
|
2016-06-13 22:20:39 +07:00
|
|
|
if (!writer_durations[i]) {
|
|
|
|
firsterr = -ENOMEM;
|
2016-01-01 09:33:22 +07:00
|
|
|
goto unwind;
|
2016-06-13 22:20:39 +07:00
|
|
|
}
|
2016-01-01 09:33:22 +07:00
|
|
|
firsterr = torture_create_kthread(rcu_perf_writer, (void *)i,
|
|
|
|
writer_tasks[i]);
|
|
|
|
if (firsterr)
|
|
|
|
goto unwind;
|
|
|
|
}
|
|
|
|
torture_init_end();
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
unwind:
|
|
|
|
torture_init_end();
|
|
|
|
rcu_perf_cleanup();
|
|
|
|
return firsterr;
|
|
|
|
}
|
|
|
|
|
|
|
|
module_init(rcu_perf_init);
|
|
|
|
module_exit(rcu_perf_cleanup);
|