New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 717347 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: Sep 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 3
Type: Bug

Blocked on:
issue 717348
issue 718379
issue 718707
issue 718717
issue 719298
issue 724392

Blocking:
issue 524758



Sign in to add a comment

Fix all the failures on the Win RandomOrder bot

Project Member Reported by tansell@chromium.org, May 2 2017

Issue description

To deploy LayoutTests on Swarming for Win, we need to fix all the tests which are currently failing.

https://luci-milo.appspot.com/buildbot/chromium.fyi/WebKit%20Win%20-%20RandomOrder/

Run locally;
------------
Total tests: 69795
* Passed: 57296 (57205 expected, 91 unexpected)
* Skipped: 11640 (11640 expected, 0 unexpected)
* Failed: 792 (792 expected, 0 unexpected)
* Flaky: 67 (67 expected, 0 unexpected)
------------

Run on swarming;
------------
Run on OS: 'Windows-7-SP1'

Total tests: 69795
* Passed: 31987 (31920 expected, 67 unexpected)
* Skipped: 36679 (36679 expected, 0 unexpected)
* Failed: 1129 (668 expected, >>>461 unexpected<<<)
* Flaky: 0 (0 expected, 0 unexpected)
------------
 
First thing to check is that we are running on the right Windows version.
Unexpected Failures:
* external/wpt/IndexedDB/interleaved-cursors.html
* http/tests/accessibility/slow-document-load.html
* http/tests/cache/history-only-cached-subresource-loads-max-age-https.html
* http/tests/cache/history-only-cached-subresource-loads.html
* http/tests/cache/loaded-from-cache-after-reload-within-iframe.html
* http/tests/cache/loaded-from-cache-after-reload.html
* http/tests/cache/location-reload.html
* http/tests/cache/preload-cleared-after-parsing-canceled-by-js.html
* http/tests/cache/subresource-expiration-1.html
* http/tests/cache/subresource-expiration-2.html
* http/tests/cache/subresource-failover-to-network.html
* http/tests/cache/subresource-multiple-instances.html
* http/tests/cookies/double-quoted-value-with-semi-colon.html
* http/tests/cookies/http-get-cookie-set-in-js.html
* http/tests/cookies/js-get-and-set-http-only-cookie.html
* http/tests/cookies/js-set-null.html
* http/tests/cookies/multiple-cookies.html
* http/tests/cookies/simple-cookies-expired.html
* http/tests/cookies/simple-cookies-max-age.html
* http/tests/cookies/single-quoted-value.html
* http/tests/encoding/meta-switch-mid-parse-with-title.html
* http/tests/encoding/meta-switch-mid-parse.html
* http/tests/fetch/serviceworker/body-mixin-base-https-other-https.html
* http/tests/fetch/serviceworker/body-mixin.html
* http/tests/fetch/serviceworker/stream-reader-base-https-other-https.html
* http/tests/fetch/serviceworker/stream-reader.html
* http/tests/fetch/window/body-mixin-base-https-other-https.html
* http/tests/fetch/window/body-mixin.html
* http/tests/fetch/window/stream-reader-base-https-other-https.html
* http/tests/fetch/window/stream-reader.html
* ... 431 more ...
Blockedon: 717348
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/4e68a5f95ca744d875773e3bb12b79ec180a5bd9

commit 4e68a5f95ca744d875773e3bb12b79ec180a5bd9
Author: tansell <tansell@chromium.org>
Date: Tue May 02 13:07:59 2017

buildbot: Force OS on the Win WebKit RandomOrder bot.

The local VM is Windows 10 and the LayoutTests are passing, so force
swarming to also run on Windows 10.

BUG= 717344 
TBR=dpranke@chromium.com,mcgreevy@chromium.org

Review-Url: https://codereview.chromium.org/2850423002
Cr-Commit-Position: refs/heads/master@{#468613}

[modify] https://crrev.com/4e68a5f95ca744d875773e3bb12b79ec180a5bd9/testing/buildbot/chromium.fyi.json

Running on Windows 10 reduced the errors down a little bit;
----------------------------------
Run on OS: 'Windows-10-10240'
swarming pending 57s
Total tests: 69840
* Passed: 30651 (30594 expected, 57 unexpected)
* Skipped: 38420 (38420 expected, 0 unexpected)
* Failed: 769 (384 expected, >>>385 unexpected<<<)
* Flaky: 0 (0 expected, 0 unexpected)
 
Unexpected Failures:
* css3/filters/blur-filter-page-scroll-parents.html
* external/wpt/IndexedDB/interleaved-cursors.html
* http/tests/accessibility/slow-document-load.html
* http/tests/cache/history-only-cached-subresource-loads-max-age-https.html
* http/tests/cache/history-only-cached-subresource-loads.html
* http/tests/cache/loaded-from-cache-after-reload-within-iframe.html
* http/tests/cache/loaded-from-cache-after-reload.html
* http/tests/cache/location-reload.html
* http/tests/cache/preload-cleared-after-parsing-canceled-by-js.html
* http/tests/cache/subresource-expiration-1.html
* http/tests/cache/subresource-expiration-2.html
* http/tests/cache/subresource-failover-to-network.html
* http/tests/cache/subresource-multiple-instances.html
* http/tests/cookies/double-quoted-value-with-semi-colon.html
* http/tests/cookies/http-get-cookie-set-in-js.html
* http/tests/cookies/js-get-and-set-http-only-cookie.html
* http/tests/cookies/js-set-null.html
* http/tests/cookies/multiple-cookies.html
* http/tests/cookies/simple-cookies-expired.html
* http/tests/cookies/simple-cookies-max-age.html
* http/tests/cookies/single-quoted-value.html
* http/tests/encoding/meta-switch-mid-parse-with-title.html
* http/tests/encoding/meta-switch-mid-parse.html
* http/tests/fetch/serviceworker/body-mixin-base-https-other-https.html
* http/tests/fetch/serviceworker/body-mixin.html
* http/tests/fetch/serviceworker/stream-reader-base-https-other-https.html
* http/tests/fetch/serviceworker/stream-reader.html
* http/tests/fetch/window/body-mixin-base-https-other-https.html
* http/tests/fetch/window/body-mixin.html
* http/tests/fetch/window/stream-reader-base-https-other-https.html
* ... 355 more ...
--------------------------------------------------

The HTTP failures appear to be caused by apache's inability to execute CGI programs;
---------------
[Wed May 03 22:57:42 2017] [error] [client 127.0.0.1] (OS 2)The system cannot find the file specified.  : couldn't create child process: 720002: subresource-failover-to-network.cgi, referer: http://127.0.0.1:8000/cache/subresource-failover-to-network.html
[Wed May 03 22:57:42 2017] [error] [client 127.0.0.1] (OS 2)The system cannot find the file specified.  : couldn't spawn child process: E:/b/swarm_slave/w/ir/third_party/WebKit/LayoutTests/http/tests/cache/resources/subresource-failover-to-network.cgi, referer: http://127.0.0.1:8000/cache/subresource-failover-to-network.html
[Wed May 03 22:57:42 2017] [error] [client 127.0.0.1] (OS 2)The system cannot find the file specified.  : couldn't create child process: 720002: random-cached.cgi, referer: http://127.0.0.1:8000/cache/loaded-from-cache-after-reload.html
[Wed May 03 22:57:42 2017] [error] [client 127.0.0.1] (OS 2)The system cannot find the file specified.  : couldn't spawn child process: E:/b/swarm_slave/w/ir/third_party/WebKit/LayoutTests/http/tests/cache/resources/random-cached.cgi, referer: http://127.0.0.1:8000/cache/loaded-from-cache-after-reload.html
---------------

What is unclear to me is why it seems to be unable to find the file specified. It is included in the isolate and other files which are side-by-side are being loaded successfully.

These cgi files are working when run on the local on the buildbot.
Owner: tansell@chromium.org
Status: Started (was: Untriaged)
Adding smut@ as he seems to understand the swarming image.

smut@ do you have any idea why the Windows Apache server when run on swarming would be unable to find these .cgi file?
It might be because on Windows we need //third_party/perl to be isolated...
Project Member

Comment 9 by bugdroid1@chromium.org, May 4 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1b05bbd2767a86f2a54b8dbf091aaf3c389842c3

commit 1b05bbd2767a86f2a54b8dbf091aaf3c389842c3
Author: tansell <tansell@chromium.org>
Date: Thu May 04 09:30:32 2017

Adding more deps to webkit-layout-tests on Windows.

 * Adding perl (needed for CGI scripts).
 * cygwin (possibly needed?)

BUG= 717347 
TBR=mcgreevy@chromium.org,dpranke@chromium.org

Review-Url: https://codereview.chromium.org/2856283003
Cr-Commit-Position: refs/heads/master@{#469297}

[modify] https://crrev.com/1b05bbd2767a86f2a54b8dbf091aaf3c389842c3/BUILD.gn

Blockedon: 718379
After adding Perl + cygwin as a dependency, the tests now pass on swarming!
--------------
webkit_layout_tests on Windows-10-10240 ( 4 mins 24 secs ) webkit_layout_tests on Windows-10-10240
Run on OS: 'Windows-10-10240'
swarming pending 16s
Total tests: 70040
* Passed: 57520 (57425 expected, 95 unexpected)
* Skipped: 11680 (11680 expected, 0 unexpected)
* Failed: 791 (791 expected, 0 unexpected)
* Flaky: 49 (49 expected, 0 unexpected)
--------------

Next would be to run the tests sharded -- however, there only seems to be one bot in the swarming pool that can do so, see  https://crbug.com/718379 
Project Member

Comment 11 by bugdroid1@chromium.org, May 4 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9fa32783bc56eb6874848a993aaca39ec5dc2591

commit 9fa32783bc56eb6874848a993aaca39ec5dc2591
Author: tansell <tansell@chromium.org>
Date: Thu May 04 11:04:34 2017

buildbot: Use Win10-10586 rather Win10-10240 on Random Order bot.

Swarming has only a single Win10-10240 bot but about 30-40 Win10-10586
bots.

TBR=dpranke@chromium.org,mcgreevy@chromium.org
BUG= 718379 , 717347 
NOTRY=true

Review-Url: https://codereview.chromium.org/2857213004
Cr-Commit-Position: refs/heads/master@{#469307}

[modify] https://crrev.com/9fa32783bc56eb6874848a993aaca39ec5dc2591/testing/buildbot/chromium.fyi.json

win-success.png
62.0 KB View Download
Project Member

Comment 13 by bugdroid1@chromium.org, May 4 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7560c4090d4fce5c16632560f76660ee9201873b

commit 7560c4090d4fce5c16632560f76660ee9201873b
Author: tansell <tansell@chromium.org>
Date: Thu May 04 13:10:57 2017

buildbot: Also run on Random Order Layout Tests on Win7-SP1

TBR=dpranke@chromium.org,mcgreevy@chromium.org
BUG= 718379 , 717347 
NOTRY=true

Review-Url: https://codereview.chromium.org/2857343003
Cr-Commit-Position: refs/heads/master@{#469318}

[modify] https://crrev.com/7560c4090d4fce5c16632560f76660ee9201873b/testing/buildbot/chromium.fyi.json

So they seem to work on Windows-10-10586
--------------
webkit_layout_tests on Windows-10-10586 webkit_layout_tests on Windows-10-10586
Run on OS: 'Windows-10-10586'
Total tests: 70042
* Passed: 57519 (57425 expected, 94 unexpected)
* Skipped: 11680 (11680 expected, 0 unexpected)
* Failed: 785 (785 expected, 0 unexpected)
* Flaky: 58 (58 expected, 0 unexpected)
--------------

And Windows-7-SP1
--------------
webkit_layout_tests webkit_layout_tests
Run on OS: 'Windows-7-SP1'
Total tests: 70042
* Passed: 57537 (57455 expected, 82 unexpected)
* Skipped: 11667 (11667 expected, 0 unexpected)
* Failed: 788 (788 expected, 0 unexpected)
* Flaky: 50 (50 expected, 0 unexpected)
--------------

So, we need to figure out which Windows we should be running on...
There are a lot more Windows-7-SP1 bots...
------------
319 bots can run this task (0 dead, 0 quarantined)
------------
Project Member

Comment 16 by bugdroid1@chromium.org, May 4 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e154b4bb6f3396b3f33d66946d1f7bf54b32e963

commit e154b4bb6f3396b3f33d66946d1f7bf54b32e963
Author: tansell <tansell@chromium.org>
Date: Thu May 04 15:14:58 2017

buildbot: Use multiple shards on Win Random Order bot.

TBR=dpranke@chromium.org,mcgreevy@chromium.org
BUG= 718379 , 717347 
NOTRY=true

Review-Url: https://codereview.chromium.org/2859173002
Cr-Commit-Position: refs/heads/master@{#469340}

[modify] https://crrev.com/e154b4bb6f3396b3f33d66946d1f7bf54b32e963/testing/buildbot/chromium.fyi.json

Turning on multiple sharding seems to have started to cause a whole bunch of failures. Currently investigating.
Project Member

Comment 18 by bugdroid1@chromium.org, May 5 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/49345d02f5e5a10cf2789d1d45b4eeec4656b542

commit 49345d02f5e5a10cf2789d1d45b4eeec4656b542
Author: tansell <tansell@chromium.org>
Date: Fri May 05 03:40:14 2017

webkitpy: Merge script fix windows directory removing.

Attempt to remove a directory tree multiple times as on Windows a
process which is currently closing could still have a file open in the
directory.

Also continue going even if we fail to remove the directory.

The merge script currently sometimes fails on Windows with the following
error;
-------------------------
Traceback (most recent call last):
  File "E:\b\c\b\WebKit_Win___RandomOrder\src\third_party\WebKit\Tools\Scripts\merge-layout-test-results", line 185, in <module>
    main(sys.argv[1:])
  File "E:\b\c\b\WebKit_Win___RandomOrder\src\third_party\WebKit\Tools\Scripts\merge-layout-test-results", line 161, in main
    shutil.rmtree(args.output_directory)
  File "E:\b\depot_tools\python276_bin\lib\shutil.py", line 256, in rmtree
    onerror(os.rmdir, path, sys.exc_info())
  File "E:\b\depot_tools\python276_bin\lib\shutil.py", line 254, in rmtree
    os.rmdir(path)
WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'E:\\b\\rr\\tmpxu2pfv\\w'
WARNING:root:merge_cmd had non-zero return code: 1
-------------------------

BUG= 717347 , 524758 

Review-Url: https://codereview.chromium.org/2862173002
Cr-Commit-Position: refs/heads/master@{#469588}

[modify] https://crrev.com/49345d02f5e5a10cf2789d1d45b4eeec4656b542/third_party/WebKit/Tools/Scripts/merge-layout-test-results

Blockedon: 718707
Blockedon: 718717
Blockedon: 719298
Project Member

Comment 22 by bugdroid1@chromium.org, May 8 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7f6264df1bd3a73c0f8af49adfc345126eebb1cd

commit 7f6264df1bd3a73c0f8af49adfc345126eebb1cd
Author: tansell <tansell@chromium.org>
Date: Mon May 08 04:56:11 2017

LayoutTests: Disable flaky tests on Windows.

These tests are failing a small amount of the time on the Windows Random
Order bots now they are running sharded on swarming. The tests affected
are;

 * fast/css-grid-layout/grid-self-baseline-two-dimensional.html
 * fast/css-grid-layout/grid-align-justify-stretch-with-orthogonal-flows.html
 * fast/dnd/dropEffect-for-effectAllowed.html
 * external/wpt/css/css-ui-3/outline-004.html

Bugs have been filed for each of the failing tests.

The most likely cause is that they depend on another test being run and
the sharding makes this no longer happening.

BUG= 717347 , 718717 , 719298 , 719299 
NOTRY=true

Review-Url: https://codereview.chromium.org/2872463002
Cr-Commit-Position: refs/heads/master@{#469907}

[modify] https://crrev.com/7f6264df1bd3a73c0f8af49adfc345126eebb1cd/third_party/WebKit/LayoutTests/TestExpectations

Comment 23 by r...@igalia.com, May 19 2017

Blockedon: 724392
Project Member

Comment 24 by bugdroid1@chromium.org, May 19 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a5c323b49440e551e321155f7a9bdb80fec5f80a

commit a5c323b49440e551e321155f7a9bdb80fec5f80a
Author: Manuel Rego Casasnovas <rego@igalia.com>
Date: Fri May 19 08:37:31 2017

LayoutTests: Disable flaky test on Windows

The following test is flacky on Windows Random Order bots:
* fast/css/font-family-case-insensitive.html

The issue is that it uses "ahem" instead of "Ahem" as font,
and for some reason that's causing the flakiness.

This patch flags it on TestExpectations.

BUG= 717347 , 724392 
TBR=qyearsley@chromium.org,tansell@chromium.org

Change-Id: I7676bfcf4ab9f6759b7a705c159621a8bfc5e496
Reviewed-on: https://chromium-review.googlesource.com/509448
Reviewed-by: Manuel Rego Casasnovas <rego@igalia.com>
Reviewed-by: Tim 'mithro' Ansell <tansell@chromium.org>
Commit-Queue: Manuel Rego Casasnovas <rego@igalia.com>
Cr-Commit-Position: refs/heads/master@{#473129}
[modify] https://crrev.com/a5c323b49440e551e321155f7a9bdb80fec5f80a/third_party/WebKit/LayoutTests/TestExpectations

Status: Fixed (was: Started)
This bot is green most of the time now!

Sign in to add a comment