mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-11-25 09:20:50 +07:00
81dc9f0ef2
In order to help benchmark the time tracepoints take, a new config option is added called CONFIG_TRACEPOINT_BENCHMARK. When this option is set a tracepoint is created called "benchmark:benchmark_event". When the tracepoint is enabled, it kicks off a kernel thread that goes into an infinite loop (calling cond_sched() to let other tasks run), and calls the tracepoint. Each iteration will record the time it took to write to the tracepoint and the next iteration that data will be passed to the tracepoint itself. That is, the tracepoint will report the time it took to do the previous tracepoint. The string written to the tracepoint is a static string of 128 bytes to keep the time the same. The initial string is simply a write of "START". The second string records the cold cache time of the first write which is not added to the rest of the calculations. As it is a tight loop, it benchmarks as hot cache. That's fine because we care most about hot paths that are probably in cache already. An example of the output: START first=3672 [COLD CACHED] last=632 first=3672 max=632 min=632 avg=316 std=446 std^2=199712 last=278 first=3672 max=632 min=278 avg=303 std=316 std^2=100337 last=277 first=3672 max=632 min=277 avg=296 std=258 std^2=67064 last=273 first=3672 max=632 min=273 avg=292 std=224 std^2=50411 last=273 first=3672 max=632 min=273 avg=288 std=200 std^2=40389 last=281 first=3672 max=632 min=273 avg=287 std=183 std^2=33666 Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
177 lines
3.7 KiB
C
177 lines
3.7 KiB
C
#include <linux/delay.h>
|
|
#include <linux/module.h>
|
|
#include <linux/kthread.h>
|
|
#include <linux/trace_clock.h>
|
|
|
|
#define CREATE_TRACE_POINTS
|
|
#include "trace_benchmark.h"
|
|
|
|
static struct task_struct *bm_event_thread;
|
|
|
|
static char bm_str[BENCHMARK_EVENT_STRLEN] = "START";
|
|
|
|
static u64 bm_total;
|
|
static u64 bm_totalsq;
|
|
static u64 bm_last;
|
|
static u64 bm_max;
|
|
static u64 bm_min;
|
|
static u64 bm_first;
|
|
static s64 bm_cnt;
|
|
|
|
/*
|
|
* This gets called in a loop recording the time it took to write
|
|
* the tracepoint. What it writes is the time statistics of the last
|
|
* tracepoint write. As there is nothing to write the first time
|
|
* it simply writes "START". As the first write is cold cache and
|
|
* the rest is hot, we save off that time in bm_first and it is
|
|
* reported as "first", which is shown in the second write to the
|
|
* tracepoint. The "first" field is writen within the statics from
|
|
* then on but never changes.
|
|
*/
|
|
static void trace_do_benchmark(void)
|
|
{
|
|
u64 start;
|
|
u64 stop;
|
|
u64 delta;
|
|
s64 stddev;
|
|
u64 seed;
|
|
u64 seedsq;
|
|
u64 last_seed;
|
|
unsigned int avg;
|
|
unsigned int std = 0;
|
|
|
|
/* Only run if the tracepoint is actually active */
|
|
if (!trace_benchmark_event_enabled())
|
|
return;
|
|
|
|
local_irq_disable();
|
|
start = trace_clock_local();
|
|
trace_benchmark_event(bm_str);
|
|
stop = trace_clock_local();
|
|
local_irq_enable();
|
|
|
|
bm_cnt++;
|
|
|
|
delta = stop - start;
|
|
|
|
/*
|
|
* The first read is cold cached, keep it separate from the
|
|
* other calculations.
|
|
*/
|
|
if (bm_cnt == 1) {
|
|
bm_first = delta;
|
|
scnprintf(bm_str, BENCHMARK_EVENT_STRLEN,
|
|
"first=%llu [COLD CACHED]", bm_first);
|
|
return;
|
|
}
|
|
|
|
bm_last = delta;
|
|
|
|
bm_total += delta;
|
|
bm_totalsq += delta * delta;
|
|
|
|
if (delta > bm_max)
|
|
bm_max = delta;
|
|
if (!bm_min || delta < bm_min)
|
|
bm_min = delta;
|
|
|
|
if (bm_cnt > 1) {
|
|
/*
|
|
* Apply Welford's method to calculate standard deviation:
|
|
* s^2 = 1 / (n * (n-1)) * (n * \Sum (x_i)^2 - (\Sum x_i)^2)
|
|
*/
|
|
stddev = (u64)bm_cnt * bm_totalsq - bm_total * bm_total;
|
|
do_div(stddev, bm_cnt);
|
|
do_div(stddev, bm_cnt - 1);
|
|
} else
|
|
stddev = 0;
|
|
|
|
delta = bm_total;
|
|
do_div(delta, bm_cnt);
|
|
avg = delta;
|
|
|
|
if (stddev > 0) {
|
|
int i = 0;
|
|
/*
|
|
* stddev is the square of standard deviation but
|
|
* we want the actualy number. Use the average
|
|
* as our seed to find the std.
|
|
*
|
|
* The next try is:
|
|
* x = (x + N/x) / 2
|
|
*
|
|
* Where N is the squared number to find the square
|
|
* root of.
|
|
*/
|
|
seed = avg;
|
|
do {
|
|
last_seed = seed;
|
|
seed = stddev;
|
|
if (!last_seed)
|
|
break;
|
|
do_div(seed, last_seed);
|
|
seed += last_seed;
|
|
do_div(seed, 2);
|
|
} while (i++ < 10 && last_seed != seed);
|
|
|
|
std = seed;
|
|
}
|
|
|
|
scnprintf(bm_str, BENCHMARK_EVENT_STRLEN,
|
|
"last=%llu first=%llu max=%llu min=%llu avg=%u std=%d std^2=%lld",
|
|
bm_last, bm_first, bm_max, bm_min, avg, std, stddev);
|
|
}
|
|
|
|
static int benchmark_event_kthread(void *arg)
|
|
{
|
|
/* sleep a bit to make sure the tracepoint gets activated */
|
|
msleep(100);
|
|
|
|
while (!kthread_should_stop()) {
|
|
|
|
trace_do_benchmark();
|
|
|
|
/*
|
|
* We don't go to sleep, but let others
|
|
* run as well.
|
|
*/
|
|
cond_resched();
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/*
|
|
* When the benchmark tracepoint is enabled, it calls this
|
|
* function and the thread that calls the tracepoint is created.
|
|
*/
|
|
void trace_benchmark_reg(void)
|
|
{
|
|
bm_event_thread = kthread_run(benchmark_event_kthread,
|
|
NULL, "event_benchmark");
|
|
WARN_ON(!bm_event_thread);
|
|
}
|
|
|
|
/*
|
|
* When the benchmark tracepoint is disabled, it calls this
|
|
* function and the thread that calls the tracepoint is deleted
|
|
* and all the numbers are reset.
|
|
*/
|
|
void trace_benchmark_unreg(void)
|
|
{
|
|
if (!bm_event_thread)
|
|
return;
|
|
|
|
kthread_stop(bm_event_thread);
|
|
|
|
strcpy(bm_str, "START");
|
|
bm_total = 0;
|
|
bm_totalsq = 0;
|
|
bm_last = 0;
|
|
bm_max = 0;
|
|
bm_min = 0;
|
|
bm_cnt = 0;
|
|
/* bm_first doesn't need to be reset but reset it anyway */
|
|
bm_first = 0;
|
|
}
|