Bug 27995

Summary:

LayoutTests are too slow

Product:

WebKit

Reporter:

Eric Seidel (no email) <eric>

Component:

Layout and Rendering

Assignee:

Nobody <webkit-unassigned>

Status:

RESOLVED CONFIGURATION CHANGED

Severity:

Normal

CC:

ahmad.saleem792, ap

Priority:

Version:

528+ (Nightly build)

Hardware:

OS:

OS X 10.5

Bug Depends on:

27998, 28008

Bug Blocks:

Attachments:

Description	Flags
Fix fast/events/click-count.html	none
slowest 1000 tests (Debug build on a Mac Pro)	none
slowest 1000 tests (Release build on a Mac Pro)	none
r47037, slowest 1000 tests (Release build on a Mac Pro)	none

Description Eric Seidel (no email) 2009-08-04 14:22:13 PDT

LayoutTests are too slow

I will post patches here to make them less slow.

Current worst 10 offenders:

9.58 secs: editing/selection/move-left-right.html
9.37 secs: fast/js/array-filter.html
7.71 secs: editing/selection/extend-selection.html
6.82 secs: fast/js/sort-randomly.html
6.25 secs: http/tests/cache/subresource-expiration.html
5.77 secs: fast/js/array-enumerators-functions.html
5.41 secs: fast/events/click-count.html
5.39 secs: fast/js/toString-and-valueOf-override.html
5.26 secs: fast/js/try-catch-crash.html
5.05 secs: http/tests/navigation/slowmetaredirect-basic.html

Comment 1 Eric Seidel (no email) 2009-08-04 14:24:21 PDT

Created attachment 34085 [details]
Fix fast/events/click-count.html

Comment 2 Eric Seidel (no email) 2009-08-04 14:44:03 PDT

Committing to http://svn.webkit.org/repository/webkit/trunk ...
	M	LayoutTests/ChangeLog
	M	LayoutTests/fast/events/click-count.html
Committed r46779
	M	LayoutTests/ChangeLog
	M	LayoutTests/fast/events/click-count.html
r46779 = e94d8b69e0213dc93223db07ffb9f9f66204ff85 (trunk)
No changes between current HEAD and refs/remotes/trunk
Resetting to the latest refs/remotes/trunk
http://trac.webkit.org/changeset/46779

Comment 3 Mark Rowe (bdash) 2009-08-04 15:08:27 PDT

Is this bug going to be used for investigating which tests are slow or for fixing individual tests?  If the latter, it's fast going to get confusing with multiple patches.  I'd suggest that you spin off separate bugs for tests that you're fixing rather than attaching them here.

Comment 4 Eric Seidel (no email) 2009-08-04 17:57:05 PDT

Created attachment 34105 [details]
slowest 1000 tests (Debug build on a Mac Pro)

Comment 5 Eric Seidel (no email) 2009-08-04 18:11:53 PDT

Comment on attachment 34085 [details]
Fix fast/events/click-count.html

@bdash: agreed.  I'll make this a meta bug.

Comment 6 Eric Seidel (no email) 2009-08-04 18:12:31 PDT

This bug only serves a temporary purpose.  It can be closed when most of the dependent bugs are closed.

Comment 7 Eric Seidel (no email) 2009-08-04 18:24:37 PDT

Created attachment 34108 [details]
slowest 1000 tests (Release build on a Mac Pro)

Comment 8 Mark Rowe (bdash) 2009-08-05 12:57:47 PDT

I sped up http/tests/misc/DOMContentLoaded-event.html and http/tests/navigation/multiple-back-forward-entries.html dramatically in r46808.

Comment 9 Eric Seidel (no email) 2009-08-11 10:13:38 PDT

Created attachment 34573 [details]
r47037, slowest 1000 tests (Release build on a Mac Pro)

Comment 10 Eric Seidel (no email) 2009-08-17 18:03:47 PDT

The commit-queue seems zippier to run these days.  I think we're making progress here!
346.82s total testing time
is my current LayoutTest time for a release build.

Comment 11 Ahmad Saleem 2022-12-30 15:26:56 PST

@ap - Do we need to track this old data anymore? I think we have now better metrics and machines and other improvement since 2009 that we can do better.

I am marking this as "RESOLVED CONFIGURATION CHANGED" (because two patches / commit landed from this) but if you think it is something still worth tracking, please share your input. Thanks!