New issue
Advanced search Search tips

Issue 605525 link

Starred by 2 users

Issue metadata

Status: WontFix
Owner: ----
Closed: Nov 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux , Windows
Pri: 3
Type: Bug



Sign in to add a comment

http/tests/xmlhttprequest/redirect-cross-origin-post.html is flaky due to upload.onload timing

Project Member Reported by kjellander@chromium.org, Apr 21 2016

Issue description

The http/tests/xmlhttprequest/redirect-cross-origin-post.html layout test is flaky on WebKit Win7 (dbg). It sometimes fails with a text diff.

I believe it started at https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%29/builds/5366/steps/webkit_tests/logs/stdio and I'm unable to figure out what to revert since the blamelist is huge.
 
Owner: dtseng@chromium.org
Status: Assigned (was: Untriaged)
dtseng@: I almost picked you at random from the blamelist. Please reassign as you see fit.
hey
Project Member

Comment 3 by bugdroid1@chromium.org, Apr 21 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/46786d3ac001458d17d3b4f897060e88746c1a0e

commit 46786d3ac001458d17d3b4f897060e88746c1a0e
Author: kjellander <kjellander@chromium.org>
Date: Thu Apr 21 14:51:20 2016

Add expectation for flaky http/tests/xmlhttprequest/redirect-cross-origin-post.html

The test is flaky since
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%29/builds/5366/steps/webkit_tests/logs/stdio
on WebKit Win7 (dbg).

BUG= 605525 
TBR=dtseng@chromium.org

Review URL: https://codereview.chromium.org/1909553003

Cr-Commit-Position: refs/heads/master@{#388772}

[modify] https://crrev.com/46786d3ac001458d17d3b4f897060e88746c1a0e/third_party/WebKit/LayoutTests/TestExpectations

Comment 4 by jianli@chromium.org, Apr 22 2016

Components: -Blink Blink>LayoutTests
Components: Blink>Network>XHR
The sequence should be for both Responding and Non-existent server:
onreadystatechange 1
onloadstart
upload.onloadstart
onreadystatechange 4
upload.onprogress
upload.onerror
onprogress
onerror

But either the sequence for Responding and Non-existent server can be:
onreadystatechange 1
onloadstart
upload.onloadstart
upload.onprogress <---
upload.onload <---
onreadystatechange 4
onprogress
onerror

https://storage.googleapis.com/chromium-layout-test-archives/WebKit_Win7__dbg_/5549/layout-test-results/http/tests/xmlhttprequest/redirect-cross-origin-post-diff.txt
https://storage.googleapis.com/chromium-layout-test-archives/WebKit_Win7__dbg_/5577/layout-test-results/http/tests/xmlhttprequest/redirect-cross-origin-post-diff.txt

Dashboard: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=http/tests/xmlhttprequest/redirect-cross-origin-post.html

Reproduced locally on Linux, very rarely though:

./third_party/WebKit/Tools/Scripts/run-webkit-tests --iterations=10000 --release http/tests/xmlhttprequest/redirect-cross-origin-post.html --exit-after-n-failures=1

About once per ~4000.
Cc: dtseng@chromium.org
Labels: OS-Linux
Owner: hirosh...@chromium.org
Status: Started (was: Assigned)
Started.

Comment 9 by sshru...@google.com, May 18 2016

Labels: Test-Layout
Components: -Blink>LayoutTests
Deprecating component:Blink>LayoutTests, to use label Test=Layout instead.
Cc: hirosh...@chromium.org
Owner: ----
Status: Available (was: Started)
Summary: http/tests/xmlhttprequest/redirect-cross-origin-post.html is flaky due to upload.onload timing (was: http/tests/xmlhttprequest/redirect-cross-origin-post.html is flaky on WebKit Win7 (dbg))
Wheather uploading is completed (and thus "upload.onload" is output) or not ("upload.onerror") before the response is received depends on timings, so the two patterns of output appear.

The test verifies the events for two XHRs are indistinguishable.
Both of the two patterns appear both of the two XHRs (not sure about the rate though), so the two XHRs are still indistinguishable, while they sometimes receive different events.

This is a problem in the test and does not reflect a real regression.

Because this is Pri-3 and a test-only issue, I left this open & Available to retain the context.
Project Member

Comment 12 by bugdroid1@chromium.org, Dec 7 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fdc595432d97064286db8e2bdda3d8eb2119dbfc

commit fdc595432d97064286db8e2bdda3d8eb2119dbfc
Author: qyearsley <qyearsley@chromium.org>
Date: Wed Dec 07 19:49:24 2016

Update TestExpectations by running Tools/Scripts/update-test-expectations.

This script removes lines for tests that appear to no longer flakily fail or timeout on the applicable platforms, according to recent results on the flakiness dashboard.

BUG=248938,446385, 513143 ,518987,522647, 532643 , 564403 , 569901 , 581468 , 605525 , 658311 , 658414 ,660308, 667953 , 671618 

Review-Url: https://codereview.chromium.org/2557593006
Cr-Commit-Position: refs/heads/master@{#437038}

[modify] https://crrev.com/fdc595432d97064286db8e2bdda3d8eb2119dbfc/third_party/WebKit/LayoutTests/TestExpectations

Project Member

Comment 13 by bugdroid1@chromium.org, Nov 1 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/be87f9d72a446ff1816cba3f7be16d771e68d5bc

commit be87f9d72a446ff1816cba3f7be16d771e68d5bc
Author: Quinten Yearsley <qyearsley@chromium.org>
Date: Wed Nov 01 15:25:46 2017

Clean up flaky expectations with update-flaky-expectations

Bug:  457273 , 605525 ,626703, 667560 ,708499, 731509 , 731535 , 732103 , 736548 , 746904 , 747163 , 752845 , 771729 
Change-Id: Ibf4be93e26a364e91d44375379aa8b5a7b59b94f
Reviewed-on: https://chromium-review.googlesource.com/736047
Reviewed-by: David Bokan <bokan@chromium.org>
Commit-Queue: Quinten Yearsley <qyearsley@chromium.org>
Cr-Commit-Position: refs/heads/master@{#513134}
[modify] https://crrev.com/be87f9d72a446ff1816cba3f7be16d771e68d5bc/third_party/WebKit/LayoutTests/TestExpectations

Status: WontFix (was: Available)
Obsolete (appears not to be flaky now)

Sign in to add a comment