skia2/infra/bots/recipes/swarm_test.py

690 lines
24 KiB
Python
Raw Normal View History

# Copyright 2016 The Chromium Authors. All rights reserved.
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.
# Recipe module for Skia Swarming test.
DEPS = [
'build/file',
'core',
'recipe_engine/json',
'recipe_engine/path',
'recipe_engine/platform',
'recipe_engine/properties',
'recipe_engine/python',
'recipe_engine/raw_io',
'flavor',
'run',
'vars',
]
TEST_BUILDERS = {
'client.skia': {
'skiabot-linux-swarm-000': [
GN: Android perf/tests Just to take inventory of which old problems still linger and which are now moot, I've gone out of my way to have this stand alone. All of gn_android_flavor's logic is self contained, without any dependency on the platform_tools scripts. The tricky bits turn out to be, copying directories containing symlinks---or really any copying involving more than one file---and getting the exit code back from adb shell. Luckily the ADB I've got on my desktop and my Nexus 5x seems to handle this all without the awkward workarounds you see here, so there's hope that One Day Soon the weird parts (basically, anything with inline python) can go away. Once we've got these bots landed green, I'll go see whether the fixes are due to ADB updates, Android updates, or perhaps something else like hardware. The parts marked TEMPORARY are a nod to the fact that the devices are used by gn_android_flavor and android_flavor both today. It's mostly about not stepping on each other's toes or leaving anything laying around that might confuse each other. The marked parts can go away when bots are either gn_ or non-gn_ but not both. I have omitted a few steps that may be important, but which are easy independent follow-ups: - running as root - locking clocks - waiting on battery levels - fancier wait-for-ready than adb wait-for-usb-device It'd be nice to, e.g., reaffirm that locking clocks helps perf stability, and that we're locking to the best policy. I've tried to keep this CL as trim as possible, leaving any of these vaguely optional steps for later. As of PS 41 or so, it looks like the trybots are all behaving as expected. We should expect no new images in Gold. Can we see trybots in Perf yet? BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2320153002 CQ_INCLUDE_TRYBOTS=master.client.skia.android:Perf-Android-Clang-AndroidOne-CPU-MT6582-arm-Debug-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Debug-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android-Trybot,Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Debug-GN_Android-Trybot,Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android-Trybot,Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Debug-GN_Android-Trybot,Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android-Trybot Review-Url: https://codereview.chromium.org/2320153002
2016-09-12 19:24:38 +00:00
'Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android',
'Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android',
'Test-Android-Clang-GalaxyS3-GPU-Mali400-arm-Debug-GN_Android',
'Test-Android-Clang-NVIDIA_Shield-GPU-TegraX1-arm64-Debug-GN_Android',
'Test-Android-Clang-Nexus10-GPU-MaliT604-arm-Release-GN_Android',
'Test-Android-Clang-Nexus6-GPU-Adreno420-arm-Debug-GN_Android',
'Test-Android-Clang-Nexus6p-GPU-Adreno430-arm64-Debug-GN_Android_Vulkan',
'Test-Android-Clang-Nexus7-GPU-Tegra3-arm-Debug-GN_Android',
'Test-Android-Clang-Nexus9-CPU-Denver-arm64-Debug-GN_Android',
'Test-Android-Clang-NexusPlayer-CPU-SSE4-x86-Release-GN_Android',
'Test-Mac-Clang-MacMini4.1-GPU-GeForce320M-x86_64-Debug',
'Test-Mac-Clang-MacMini6.2-CPU-AVX-x86_64-Debug',
'Test-Mac-Clang-MacMini6.2-GPU-HD4000-x86_64-Debug-CommandBuffer',
'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86-Debug',
'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Debug',
'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Debug-ASAN',
'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Debug-MSAN',
'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Shared',
'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-TSAN',
'Test-Ubuntu-GCC-ShuttleA-GPU-GTX550Ti-x86_64-Release-Valgrind',
'Test-Win10-MSVC-ShuttleA-GPU-GTX660-x86_64-Debug-Vulkan',
'Test-Win8-MSVC-ShuttleB-CPU-AVX2-x86_64-Release-Trybot',
'Test-Win8-MSVC-ShuttleB-GPU-GTX960-x86_64-Debug-ANGLE',
GN: Android perf/tests Just to take inventory of which old problems still linger and which are now moot, I've gone out of my way to have this stand alone. All of gn_android_flavor's logic is self contained, without any dependency on the platform_tools scripts. The tricky bits turn out to be, copying directories containing symlinks---or really any copying involving more than one file---and getting the exit code back from adb shell. Luckily the ADB I've got on my desktop and my Nexus 5x seems to handle this all without the awkward workarounds you see here, so there's hope that One Day Soon the weird parts (basically, anything with inline python) can go away. Once we've got these bots landed green, I'll go see whether the fixes are due to ADB updates, Android updates, or perhaps something else like hardware. The parts marked TEMPORARY are a nod to the fact that the devices are used by gn_android_flavor and android_flavor both today. It's mostly about not stepping on each other's toes or leaving anything laying around that might confuse each other. The marked parts can go away when bots are either gn_ or non-gn_ but not both. I have omitted a few steps that may be important, but which are easy independent follow-ups: - running as root - locking clocks - waiting on battery levels - fancier wait-for-ready than adb wait-for-usb-device It'd be nice to, e.g., reaffirm that locking clocks helps perf stability, and that we're locking to the best policy. I've tried to keep this CL as trim as possible, leaving any of these vaguely optional steps for later. As of PS 41 or so, it looks like the trybots are all behaving as expected. We should expect no new images in Gold. Can we see trybots in Perf yet? BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2320153002 CQ_INCLUDE_TRYBOTS=master.client.skia.android:Perf-Android-Clang-AndroidOne-CPU-MT6582-arm-Debug-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Debug-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android-Trybot,Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Debug-GN_Android-Trybot,Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android-Trybot,Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Debug-GN_Android-Trybot,Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android-Trybot Review-Url: https://codereview.chromium.org/2320153002
2016-09-12 19:24:38 +00:00
'Test-iOS-Clang-iPad4-GPU-SGX554-Arm7-Debug',
],
},
}
def dm_flags(bot):
args = []
# 32-bit desktop bots tend to run out of memory, because they have relatively
# far more cores than RAM (e.g. 32 cores, 3G RAM). Hold them back a bit.
if '-x86-' in bot and not 'NexusPlayer' in bot:
args.extend('--threads 4'.split(' '))
# These are the canonical configs that we would ideally run on all bots. We
# may opt out or substitute some below for specific bots
configs = ['8888', 'gpu', 'gpusrgb', 'pdf']
# Add in either msaa4 or msaa16 to the canonical set of configs to run
if 'Android' in bot or 'iOS' in bot:
configs.append('msaa4')
else:
configs.append('msaa16')
# With msaa, the S4 crashes and the NP produces a long error stream when we
# run with MSAA. The Tegra2 and Tegra3 just don't support it. No record of
# why we're not running msaa on iOS, probably started with gpu config and just
# haven't tried.
if ('GalaxyS4' in bot or
'NexusPlayer' in bot or
'Tegra3' in bot or
'iOS' in bot or
# skia:5792
'iHD530' in bot):
configs = [x for x in configs if 'msaa' not in x]
# Runs out of memory on Android bots and Daisy. Everyone else seems fine.
if 'Android' in bot or 'Daisy' in bot:
configs.remove('pdf')
if '-GCE-' in bot:
configs.extend(['565'])
configs.extend(['f16', 'srgb']) # Gamma-correct formats.
configs.extend(['sp-8888', '2ndpic-8888']) # Test niche uses of SkPicture.
configs.extend(['lite-8888']) # Experimental display list.
if '-TSAN' not in bot:
if ('TegraK1' in bot or
'TegraX1' in bot or
'GTX550Ti' in bot or
'GTX660' in bot or
'GT610' in bot):
if 'Android' in bot:
configs.append('nvprdit4')
else:
configs.append('nvprdit16')
# We want to test the OpenGL config not the GLES config on the X1
if 'TegraX1' in bot:
configs = [x.replace('gpu', 'gl') for x in configs]
configs = [x.replace('msaa', 'glmsaa') for x in configs]
configs = [x.replace('nvpr', 'glnvpr') for x in configs]
# NP is running out of RAM when we run all these modes. skia:3255
if 'NexusPlayer' not in bot:
configs.extend(mode + '-8888' for mode in
['serialize', 'tiles_rt', 'pic'])
if 'ANGLE' in bot:
configs.extend(['angle_d3d11_es2',
'angle_d3d9_es2',
'angle_d3d11_es2_msaa4'])
# We want to run gpudft on atleast the mali 400
if 'GalaxyS3' in bot:
configs.append('gpudft')
# Test instanced rendering on a limited number of platforms
if 'Nexus6' in bot:
configs.append('esinst') # esinst4 isn't working yet on Adreno.
elif 'TegraX1' in bot:
# Multisampled instanced configs use nvpr.
configs = [x.replace('glnvpr', 'glinst') for x in configs]
configs.append('glinst')
elif 'MacMini6.2' in bot:
configs.extend(['glinst', 'glinst16'])
# CommandBuffer bot *only* runs the command_buffer config.
if 'CommandBuffer' in bot:
configs = ['commandbuffer']
# Vulkan bot *only* runs the vk config.
if 'Vulkan' in bot:
configs = ['vk']
args.append('--config')
args.extend(configs)
# Run tests, gms, and image decoding tests everywhere.
args.extend('--src tests gm image colorImage svg'.split(' '))
if 'GalaxyS' in bot:
args.extend(('--threads', '0'))
blacklisted = []
def blacklist(quad):
config, src, options, name = quad.split(' ') if type(quad) is str else quad
if config == '_' or config in configs:
blacklisted.extend([config, src, options, name])
# TODO: ???
blacklist('f16 _ _ dstreadshuffle')
blacklist('f16 image _ _')
blacklist('srgb image _ _')
blacklist('gpusrgb image _ _')
blacklist('glsrgb image _ _')
if 'Valgrind' in bot:
# These take 18+ hours to run.
blacklist('pdf gm _ fontmgr_iter')
blacklist('pdf _ _ PANO_20121023_214540.jpg')
blacklist('pdf skp _ worldjournal')
blacklist('pdf skp _ desk_baidu.skp')
blacklist('pdf skp _ desk_wikipedia.skp')
blacklist('_ svg _ _')
if 'iOS' in bot:
blacklist('gpu skp _ _')
blacklist('msaa skp _ _')
blacklist('msaa16 gm _ tilemodesProcess')
if 'Mac' in bot or 'iOS' in bot:
# CG fails on questionable bmps
blacklist('_ image gen_platf rgba32abf.bmp')
blacklist('_ image gen_platf rgb24prof.bmp')
blacklist('_ image gen_platf rgb24lprof.bmp')
blacklist('_ image gen_platf 8bpp-pixeldata-cropped.bmp')
blacklist('_ image gen_platf 4bpp-pixeldata-cropped.bmp')
blacklist('_ image gen_platf 32bpp-pixeldata-cropped.bmp')
blacklist('_ image gen_platf 24bpp-pixeldata-cropped.bmp')
# CG has unpredictable behavior on this questionable gif
# It's probably using uninitialized memory
blacklist('_ image gen_platf frame_larger_than_image.gif')
# CG has unpredictable behavior on incomplete pngs
# skbug.com/5774
blacklist('_ image gen_platf inc0.png')
blacklist('_ image gen_platf inc1.png')
blacklist('_ image gen_platf inc2.png')
blacklist('_ image gen_platf inc3.png')
blacklist('_ image gen_platf inc4.png')
blacklist('_ image gen_platf inc5.png')
blacklist('_ image gen_platf inc6.png')
blacklist('_ image gen_platf inc7.png')
blacklist('_ image gen_platf inc8.png')
blacklist('_ image gen_platf inc9.png')
blacklist('_ image gen_platf inc10.png')
blacklist('_ image gen_platf inc11.png')
blacklist('_ image gen_platf inc12.png')
blacklist('_ image gen_platf inc13.png')
blacklist('_ image gen_platf inc14.png')
# WIC fails on questionable bmps
if 'Win' in bot:
blacklist('_ image gen_platf rle8-height-negative.bmp')
blacklist('_ image gen_platf rle4-height-negative.bmp')
blacklist('_ image gen_platf pal8os2v2.bmp')
blacklist('_ image gen_platf pal8os2v2-16.bmp')
blacklist('_ image gen_platf rgba32abf.bmp')
blacklist('_ image gen_platf rgb24prof.bmp')
blacklist('_ image gen_platf rgb24lprof.bmp')
blacklist('_ image gen_platf 8bpp-pixeldata-cropped.bmp')
blacklist('_ image gen_platf 4bpp-pixeldata-cropped.bmp')
blacklist('_ image gen_platf 32bpp-pixeldata-cropped.bmp')
blacklist('_ image gen_platf 24bpp-pixeldata-cropped.bmp')
if 'x86_64' in bot and 'CPU' in bot:
# This GM triggers a SkSmallAllocator assert.
blacklist('_ gm _ composeshader_bitmap')
if 'Android' in bot or 'iOS' in bot:
# This test crashes the N9 (perhaps because of large malloc/frees). It also
# is fairly slow and not platform-specific. So we just disable it on all of
# Android and iOS. skia:5438
blacklist('_ test _ GrShape')
if 'Win8' in bot:
# bungeman: "Doesn't work on Windows anyway, produces unstable GMs with
# 'Unexpected error' from DirectWrite"
blacklist('_ gm _ fontscalerdistortable')
# skia:5636
blacklist('_ svg _ Nebraska-StateSeal.svg')
# skia:4095
bad_serialize_gms = ['bleed_image',
'c_gms',
'colortype',
'colortype_xfermodes',
'drawfilter',
'fontmgr_bounds_0.75_0',
'fontmgr_bounds_1_-0.25',
'fontmgr_bounds',
'fontmgr_match',
'fontmgr_iter',
'imagemasksubset']
# skia:5589
bad_serialize_gms.extend(['bitmapfilters',
'bitmapshaders',
'bleed',
'bleed_alpha_bmp',
'bleed_alpha_bmp_shader',
'convex_poly_clip',
'extractalpha',
'filterbitmap_checkerboard_32_32_g8',
'filterbitmap_image_mandrill_64',
'shadows',
'simpleaaclip_aaclip'])
# skia:5595
bad_serialize_gms.extend(['composeshader_bitmap',
'scaled_tilemodes_npot',
'scaled_tilemodes'])
# skia:5778
bad_serialize_gms.append('typefacerendering_pfaMac')
for test in bad_serialize_gms:
blacklist(['serialize-8888', 'gm', '_', test])
if 'Mac' not in bot:
for test in ['bleed_alpha_image', 'bleed_alpha_image_shader']:
blacklist(['serialize-8888', 'gm', '_', test])
# It looks like we skip these only for out-of-memory concerns.
if 'Win' in bot or 'Android' in bot:
for test in ['verylargebitmap', 'verylarge_picture_image']:
blacklist(['serialize-8888', 'gm', '_', test])
# skia:4769
for test in ['drawfilter']:
blacklist([ 'sp-8888', 'gm', '_', test])
blacklist([ 'pic-8888', 'gm', '_', test])
blacklist(['2ndpic-8888', 'gm', '_', test])
blacklist([ 'lite-8888', 'gm', '_', test])
# skia:4703
for test in ['image-cacherator-from-picture',
'image-cacherator-from-raster',
'image-cacherator-from-ctable']:
blacklist([ 'sp-8888', 'gm', '_', test])
blacklist([ 'pic-8888', 'gm', '_', test])
blacklist([ '2ndpic-8888', 'gm', '_', test])
blacklist(['serialize-8888', 'gm', '_', test])
# GM that requires raster-backed canvas
for test in ['gamut']:
blacklist([ 'sp-8888', 'gm', '_', test])
blacklist([ 'pic-8888', 'gm', '_', test])
blacklist([ 'lite-8888', 'gm', '_', test])
blacklist([ '2ndpic-8888', 'gm', '_', test])
blacklist(['serialize-8888', 'gm', '_', test])
# Extensions for RAW images
r = ["arw", "cr2", "dng", "nef", "nrw", "orf", "raf", "rw2", "pef", "srw",
"ARW", "CR2", "DNG", "NEF", "NRW", "ORF", "RAF", "RW2", "PEF", "SRW"]
# skbug.com/4888
# Blacklist RAW images (and a few large PNGs) on GPU bots
# until we can resolve failures
if 'GPU' in bot:
blacklist('_ image _ interlaced1.png')
blacklist('_ image _ interlaced2.png')
blacklist('_ image _ interlaced3.png')
for raw_ext in r:
blacklist('_ image _ .%s' % raw_ext)
if 'Nexus9' in bot:
for raw_ext in r:
blacklist('_ image _ .%s' % raw_ext)
# Large image that overwhelms older Mac bots
if 'MacMini4.1-GPU' in bot:
blacklist('_ image _ abnormal.wbmp')
blacklist(['msaa16', 'gm', '_', 'blurcircles'])
match = []
if 'Valgrind' in bot: # skia:3021
match.append('~Threaded')
if 'GalaxyS3' in bot: # skia:1699
match.append('~WritePixels')
if 'AndroidOne' in bot: # skia:4711
match.append('~WritePixels')
if 'NexusPlayer' in bot:
match.append('~ResourceCache')
if 'Nexus10' in bot: # skia:5509
match.append('~CopySurface')
if 'ANGLE' in bot and 'Debug' in bot:
match.append('~GLPrograms') # skia:4717
if 'MSAN' in bot:
match.extend(['~Once', '~Shared']) # Not sure what's up with these tests.
if 'TSAN' in bot:
match.extend(['~ReadWriteAlpha']) # Flaky on TSAN-covered on nvidia bots.
if 'Vulkan' in bot and 'Adreno' in bot:
# skia:5777
match.extend(['~XfermodeImageFilterCroppedInput',
'~GrTextureStripAtlasFlush',
'~CopySurface'])
if blacklisted:
args.append('--blacklist')
args.extend(blacklisted)
if match:
args.append('--match')
args.extend(match)
# These bots run out of memory running RAW codec tests. Do not run them in
# parallel
if ('NexusPlayer' in bot or 'Nexus5' in bot or 'Nexus9' in bot
or 'Win8-MSVC-ShuttleB' in bot):
args.append('--noRAW_threading')
return args
def key_params(api):
"""Build a unique key from the builder name (as a list).
E.g. arch x86 gpu GeForce320M mode MacMini4.1 os Mac10.6
"""
# Don't bother to include role, which is always Test.
# TryBots are uploaded elsewhere so they can use the same key.
blacklist = ['role', 'is_trybot']
flat = []
for k in sorted(api.vars.builder_cfg.keys()):
if k not in blacklist:
flat.append(k)
flat.append(api.vars.builder_cfg[k])
return flat
def test_steps(api):
"""Run the DM test."""
use_hash_file = False
if api.vars.upload_dm_results:
# This must run before we write anything into
# api.flavor.device_dirs.dm_dir or we may end up deleting our
# output on machines where they're the same.
api.flavor.create_clean_host_dir(api.vars.dm_dir)
host_dm_dir = str(api.vars.dm_dir)
device_dm_dir = str(api.flavor.device_dirs.dm_dir)
if host_dm_dir != device_dm_dir:
api.flavor.create_clean_device_dir(device_dm_dir)
# Obtain the list of already-generated hashes.
hash_filename = 'uninteresting_hashes.txt'
# Ensure that the tmp_dir exists.
api.run.run_once(api.file.makedirs,
'tmp_dir',
api.vars.tmp_dir,
infra_step=True)
host_hashes_file = api.vars.tmp_dir.join(hash_filename)
hashes_file = api.flavor.device_path_join(
api.flavor.device_dirs.tmp_dir, hash_filename)
api.run(
api.python.inline,
'get uninteresting hashes',
program="""
import contextlib
import math
import socket
import sys
import time
import urllib2
HASHES_URL = 'https://gold.skia.org/_/hashes'
RETRIES = 5
TIMEOUT = 60
WAIT_BASE = 15
socket.setdefaulttimeout(TIMEOUT)
for retry in range(RETRIES):
try:
with contextlib.closing(
urllib2.urlopen(HASHES_URL, timeout=TIMEOUT)) as w:
hashes = w.read()
with open(sys.argv[1], 'w') as f:
f.write(hashes)
break
except Exception as e:
print 'Failed to get uninteresting hashes from %s:' % HASHES_URL
print e
if retry == RETRIES:
raise
waittime = WAIT_BASE * math.pow(2, retry)
print 'Retry in %d seconds.' % waittime
time.sleep(waittime)
""",
args=[host_hashes_file],
cwd=api.vars.skia_dir,
abort_on_failure=False,
fail_build_on_failure=False,
infra_step=True)
if api.path.exists(host_hashes_file):
api.flavor.copy_file_to_device(host_hashes_file, hashes_file)
use_hash_file = True
# Run DM.
properties = [
'gitHash', api.vars.got_revision,
'master', api.vars.master_name,
'builder', api.vars.builder_name,
'build_number', api.vars.build_number,
]
if api.vars.is_trybot:
properties.extend([
'issue', api.vars.issue,
'patchset', api.vars.patchset,
'patch_storage', api.vars.patch_storage,
])
args = [
'dm',
'--undefok', # This helps branches that may not know new flags.
'--resourcePath', api.flavor.device_dirs.resource_dir,
'--skps', api.flavor.device_dirs.skp_dir,
'--images', api.flavor.device_path_join(
api.flavor.device_dirs.images_dir, 'dm'),
'--colorImages', api.flavor.device_path_join(
api.flavor.device_dirs.images_dir, 'colorspace'),
'--nameByHash',
'--properties'
] + properties
args.extend(['--svgs', api.flavor.device_dirs.svg_dir])
args.append('--key')
args.extend(key_params(api))
if use_hash_file:
args.extend(['--uninterestingHashesFile', hashes_file])
if api.vars.upload_dm_results:
args.extend(['--writePath', api.flavor.device_dirs.dm_dir])
skip_flag = None
if api.vars.builder_cfg.get('cpu_or_gpu') == 'CPU':
skip_flag = '--nogpu'
elif api.vars.builder_cfg.get('cpu_or_gpu') == 'GPU':
skip_flag = '--nocpu'
if skip_flag:
args.append(skip_flag)
args.extend(dm_flags(api.vars.builder_name))
api.run(api.flavor.step, 'dm', cmd=args,
abort_on_failure=False,
env=api.vars.default_env)
if api.vars.upload_dm_results:
# Copy images and JSON to host machine if needed.
api.flavor.copy_directory_contents_to_host(
api.flavor.device_dirs.dm_dir, api.vars.dm_dir)
# See skia:2789.
if ('Valgrind' in api.vars.builder_name and
api.vars.builder_cfg.get('cpu_or_gpu') == 'GPU'):
abandonGpuContext = list(args)
abandonGpuContext.append('--abandonGpuContext')
api.run(api.flavor.step, 'dm --abandonGpuContext',
cmd=abandonGpuContext, abort_on_failure=False)
preAbandonGpuContext = list(args)
preAbandonGpuContext.append('--preAbandonGpuContext')
api.run(api.flavor.step, 'dm --preAbandonGpuContext',
cmd=preAbandonGpuContext, abort_on_failure=False,
env=api.vars.default_env)
def RunSteps(api):
api.core.setup()
GN: Android perf/tests Just to take inventory of which old problems still linger and which are now moot, I've gone out of my way to have this stand alone. All of gn_android_flavor's logic is self contained, without any dependency on the platform_tools scripts. The tricky bits turn out to be, copying directories containing symlinks---or really any copying involving more than one file---and getting the exit code back from adb shell. Luckily the ADB I've got on my desktop and my Nexus 5x seems to handle this all without the awkward workarounds you see here, so there's hope that One Day Soon the weird parts (basically, anything with inline python) can go away. Once we've got these bots landed green, I'll go see whether the fixes are due to ADB updates, Android updates, or perhaps something else like hardware. The parts marked TEMPORARY are a nod to the fact that the devices are used by gn_android_flavor and android_flavor both today. It's mostly about not stepping on each other's toes or leaving anything laying around that might confuse each other. The marked parts can go away when bots are either gn_ or non-gn_ but not both. I have omitted a few steps that may be important, but which are easy independent follow-ups: - running as root - locking clocks - waiting on battery levels - fancier wait-for-ready than adb wait-for-usb-device It'd be nice to, e.g., reaffirm that locking clocks helps perf stability, and that we're locking to the best policy. I've tried to keep this CL as trim as possible, leaving any of these vaguely optional steps for later. As of PS 41 or so, it looks like the trybots are all behaving as expected. We should expect no new images in Gold. Can we see trybots in Perf yet? BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2320153002 CQ_INCLUDE_TRYBOTS=master.client.skia.android:Perf-Android-Clang-AndroidOne-CPU-MT6582-arm-Debug-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Debug-GN_Android-Trybot,Perf-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android-Trybot,Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Debug-GN_Android-Trybot,Test-Android-Clang-AndroidOne-CPU-MT6582-arm-Release-GN_Android-Trybot,Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Debug-GN_Android-Trybot,Test-Android-Clang-AndroidOne-GPU-Mali400MP2-arm-Release-GN_Android-Trybot Review-Url: https://codereview.chromium.org/2320153002
2016-09-12 19:24:38 +00:00
try:
api.flavor.install()
test_steps(api)
finally:
api.flavor.cleanup_steps()
api.run.check_failure()
def GenTests(api):
for mastername, slaves in TEST_BUILDERS.iteritems():
for slavename, builders_by_slave in slaves.iteritems():
for builder in builders_by_slave:
test = (
api.test(builder) +
api.properties(buildername=builder,
mastername=mastername,
slavename=slavename,
buildnumber=5,
revision='abc123',
path_config='kitchen',
swarm_out_dir='[SWARM_OUT_DIR]') +
api.path.exists(
api.path['slave_build'].join('skia'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skimage', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skp', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'svg', 'VERSION'),
api.path['slave_build'].join('tmp', 'uninteresting_hashes.txt')
)
)
if 'Trybot' in builder:
test += api.properties(issue=500,
patchset=1,
rietveld='https://codereview.chromium.org')
if 'Win' in builder:
test += api.platform('win', 64)
yield test
builder = 'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Debug'
yield (
api.test('failed_dm') +
api.properties(buildername=builder,
mastername='client.skia',
slavename='skiabot-linux-swarm-000',
buildnumber=6,
revision='abc123',
path_config='kitchen',
swarm_out_dir='[SWARM_OUT_DIR]') +
api.path.exists(
api.path['slave_build'].join('skia'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skimage', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skp', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'svg', 'VERSION'),
api.path['slave_build'].join('tmp', 'uninteresting_hashes.txt')
) +
api.step_data('dm', retcode=1)
)
builder = 'Test-Android-Clang-Nexus7-GPU-Tegra3-arm-Debug-GN_Android'
yield (
api.test('failed_get_hashes') +
api.properties(buildername=builder,
mastername='client.skia',
slavename='skiabot-linux-swarm-000',
buildnumber=6,
revision='abc123',
path_config='kitchen',
swarm_out_dir='[SWARM_OUT_DIR]') +
api.path.exists(
api.path['slave_build'].join('skia'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skimage', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skp', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'svg', 'VERSION'),
api.path['slave_build'].join('tmp', 'uninteresting_hashes.txt')
) +
api.step_data('get uninteresting hashes', retcode=1)
)
builder = 'Test-iOS-Clang-iPad4-GPU-SGX554-Arm7-Debug'
yield (
api.test('missing_SKP_VERSION_device') +
api.properties(buildername=builder,
mastername='client.skia',
slavename='skiabot-linux-swarm-000',
buildnumber=6,
revision='abc123',
path_config='kitchen',
swarm_out_dir='[SWARM_OUT_DIR]') +
api.path.exists(
api.path['slave_build'].join('skia'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skimage', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skp', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'svg', 'VERSION'),
api.path['slave_build'].join('tmp', 'uninteresting_hashes.txt')
) +
api.step_data('read SKP_VERSION', retcode=1)
)
builder = 'Test-Win8-MSVC-ShuttleB-CPU-AVX2-x86_64-Release-Trybot'
yield (
api.test('big_issue_number') +
api.properties(buildername=builder,
mastername='client.skia.compile',
slavename='skiabot-linux-swarm-000',
buildnumber=5,
revision='abc123',
path_config='kitchen',
swarm_out_dir='[SWARM_OUT_DIR]',
rietveld='https://codereview.chromium.org',
patchset=1,
issue=2147533002L) +
api.path.exists(
api.path['slave_build'].join('skia'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skimage', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'skp', 'VERSION'),
api.path['slave_build'].join('skia', 'infra', 'bots', 'assets',
'svg', 'VERSION'),
api.path['slave_build'].join('tmp', 'uninteresting_hashes.txt')
) +
api.platform('win', 64)
)
builder = 'Test-Ubuntu-GCC-GCE-CPU-AVX2-x86-Debug-Trybot'
gerrit_kwargs = {
'patch_storage': 'gerrit',
'repository': 'skia',
'event.patchSet.ref': 'refs/changes/00/2100/2',
'event.change.number': '2100',
}
yield (
api.test('recipe_with_gerrit_patch') +
api.properties(
buildername=builder,
mastername='client.skia',
slavename='skiabot-linux-swarm-000',
buildnumber=5,
path_config='kitchen',
swarm_out_dir='[SWARM_OUT_DIR]',
revision='abc123',
**gerrit_kwargs)
)