tools/profiling/microbenchmarks/bm_diff.py

#!/usr/bin/env python2.7
# Copyright 2017, Google Inc.
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
#     * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
#     * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following disclaimer
# in the documentation and/or other materials provided with the
# distribution.
#     * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

import sys
import json
import bm_json
import tabulate
import argparse
from scipy import stats
import subprocess
import multiprocessing
import collections
import pipes
import os
sys.path.append(os.path.join(os.path.dirname(sys.argv[0]), '..', '..', 'run_tests', 'python_utils'))
import comment_on_pr
import jobset
import itertools
import speedup
import random
import shutil
import errno

_INTERESTING = (
  'cpu_time',
  'real_time',
  'locks_per_iteration',
  'allocs_per_iteration',
  'writes_per_iteration',
  'atm_cas_per_iteration',
  'atm_add_per_iteration',
)

def changed_ratio(n, o):
  if float(o) <= .0001: o = 0
  if float(n) <= .0001: n = 0
  if o == 0 and n == 0: return 0
  if o == 0: return 100
  return (float(n)-float(o))/float(o)

def median(ary):
  ary = sorted(ary)
  n = len(ary)
  if n%2 == 0:
    return (ary[n/2] + ary[n/2+1]) / 2.0
  else:
    return ary[n/2]

def min_change(pct):
  return lambda n, o: abs(changed_ratio(n,o)) > pct/100.0

_AVAILABLE_BENCHMARK_TESTS = ['bm_fullstack_unary_ping_pong',
                              'bm_fullstack_streaming_ping_pong',
                              'bm_fullstack_streaming_pump',
                              'bm_closure',
                              'bm_cq',
                              'bm_call_create',
                              'bm_error',
                              'bm_chttp2_hpack',
                              'bm_chttp2_transport',
                              'bm_pollset',
                              'bm_metadata',
                              'bm_fullstack_trickle']

argp = argparse.ArgumentParser(description='Perform diff on microbenchmarks')
argp.add_argument('-t', '--track',
                  choices=sorted(_INTERESTING),
                  nargs='+',
                  default=sorted(_INTERESTING),
                  help='Which metrics to track')
argp.add_argument('-b', '--benchmarks', nargs='+', choices=_AVAILABLE_BENCHMARK_TESTS, default=['bm_cq'])
argp.add_argument('-d', '--diff_base', type=str)
argp.add_argument('-r', '--repetitions', type=int, default=1)
argp.add_argument('-l', '--loops', type=int, default=12)
argp.add_argument('-j', '--jobs', type=int, default=multiprocessing.cpu_count())
args = argp.parse_args()

assert args.diff_base

def avg(lst):
  sum = 0.0
  n = 0.0
  for el in lst:
    sum += el
    n += 1
  return sum / n

def make_cmd(cfg):
  return ['make'] + args.benchmarks + [
      'CONFIG=%s' % cfg, '-j', '%d' % args.jobs]

def build(dest):
  shutil.rmtree('bm_diff_%s' % dest, ignore_errors=True)
  subprocess.check_call(['git', 'submodule', 'update'])
  try:
    subprocess.check_call(make_cmd('opt'))
    subprocess.check_call(make_cmd('counters'))
  except subprocess.CalledProcessError, e:
    subprocess.check_call(['make', 'clean'])
    subprocess.check_call(make_cmd('opt'))
    subprocess.check_call(make_cmd('counters'))
  os.rename('bins', 'bm_diff_%s' % dest)

def collect1(bm, cfg, ver, idx):
  cmd = ['bm_diff_%s/%s/%s' % (ver, cfg, bm),
         '--benchmark_out=%s.%s.%s.%d.json' % (bm, cfg, ver, idx),
         '--benchmark_out_format=json',
         '--benchmark_repetitions=%d' % (args.repetitions)
         ]
  return jobset.JobSpec(cmd, shortname='%s %s %s %d/%d' % (bm, cfg, ver, idx+1, args.loops),
                             verbose_success=True, timeout_seconds=None)

build('new')

where_am_i = subprocess.check_output(['git', 'rev-parse', '--abbrev-ref', 'HEAD']).strip()
subprocess.check_call(['git', 'checkout', args.diff_base])
try:
  build('old')
finally:
  subprocess.check_call(['git', 'checkout', where_am_i])
  subprocess.check_call(['git', 'submodule', 'update'])

jobs = []
for loop in range(0, args.loops):
  jobs.extend(x for x in itertools.chain(
    (collect1(bm, 'opt', 'new', loop) for bm in args.benchmarks),
    (collect1(bm, 'counters', 'new', loop) for bm in args.benchmarks),
    (collect1(bm, 'opt', 'old', loop) for bm in args.benchmarks),
    (collect1(bm, 'counters', 'old', loop) for bm in args.benchmarks),
  ))
random.shuffle(jobs, random.SystemRandom().random)

jobset.run(jobs, maxjobs=args.jobs)

class Benchmark:

  def __init__(self):
    self.samples = {
      True: collections.defaultdict(list),
      False: collections.defaultdict(list)
    }
    self.final = {}

  def add_sample(self, data, new):
    for f in args.track:
      if f in data:
        self.samples[new][f].append(float(data[f]))

  def process(self):
    for f in sorted(args.track):
      new = self.samples[True][f]
      old = self.samples[False][f]
      if not new or not old: continue
      mdn_diff = abs(median(new) - median(old))
      print '%s: new=%r old=%r mdn_diff=%r' % (f, new, old, mdn_diff)
      s = speedup.speedup(new, old)
      if abs(s) > 3 and mdn_diff > 0.5:
        self.final[f] = '%+d%%' % s
    return self.final.keys()

  def skip(self):
    return not self.final

  def row(self, flds):
    return [self.final[f] if f in self.final else '' for f in flds]


def eintr_be_gone(fn):
  """Run fn until it doesn't stop because of EINTR"""
  while True:
    try:
      return fn()
    except IOError, e:
      if e.errno != errno.EINTR:
        raise


def read_json(filename):
  with open(filename) as f: return json.loads(f.read())


def finalize():
  benchmarks = collections.defaultdict(Benchmark)

  for bm in args.benchmarks:
    for loop in range(0, args.loops):
      js_new_ctr = read_json('%s.counters.new.%d.json' % (bm, loop))
      js_new_opt = read_json('%s.opt.new.%d.json' % (bm, loop))
      js_old_ctr = read_json('%s.counters.old.%d.json' % (bm, loop))
      js_old_opt = read_json('%s.opt.old.%d.json' % (bm, loop))

      for row in bm_json.expand_json(js_new_ctr, js_new_opt):
        print row
        name = row['cpp_name']
        if name.endswith('_mean') or name.endswith('_stddev'): continue
        benchmarks[name].add_sample(row, True)
      for row in bm_json.expand_json(js_old_ctr, js_old_opt):
        print row
        name = row['cpp_name']
        if name.endswith('_mean') or name.endswith('_stddev'): continue
        benchmarks[name].add_sample(row, False)

  really_interesting = set()
  for name, bm in benchmarks.items():
    print name
    really_interesting.update(bm.process())
  fields = [f for f in args.track if f in really_interesting]

  headers = ['Benchmark'] + fields
  rows = []
  for name in sorted(benchmarks.keys()):
    if benchmarks[name].skip(): continue
    rows.append([name] + benchmarks[name].row(fields))
  if rows:
    text = 'Performance differences noted:\n' + tabulate.tabulate(rows, headers=headers, floatfmt='+.2f')
  else:
    text = 'No significant performance differences'
  comment_on_pr.comment_on_pr('```\n%s\n```' % text)
  print text


eintr_be_gone(finalize)