Bug 80441 – Avoid unnecessary work when setting attributes to identical value.

• Component: DOM
• Status: RESOLVED LATER

[Andreas Kling]

We should see if we can avoid the whole attributeChanged() chain when setting an element attribute to the same value.

[Andreas Kling]

Created attachment 130433 [details]
Patch

Let's see how EWS feels about this.

[Andreas Kling]

There are some failures from the patch as-is.

- For some editing tests, we depend on the style recalc triggered by setting the "style" attribute, even if to the same value.
- There also appear to be some tests depending on network reloads being triggered by re-setting of the src attribute.

Needs more investigation.

[Darin Adler]

(In reply to comment #2)
> - For some editing tests, we depend on the style recalc triggered by setting the "style" attribute, even if to the same value.

Seems fixable.

> - There also appear to be some tests depending on network reloads being triggered by re-setting of the src attribute.

Seems like a fundamental design decision, so maybe not fixable.

[Darin Adler]

(In reply to comment #3)
> (In reply to comment #2)
> > - There also appear to be some tests depending on network reloads being triggered by re-setting of the src attribute.
> 
> Seems like a fundamental design decision, so maybe not fixable.

I mean maybe not fixable without some special status for attributes that need this sort of behavior.

[Andreas Kling]

Created attachment 199500 [details]
Lunch for EWS

Let's see what EWS thinks about this today.

[Build Bot]

Comment on attachment 199500 [details]
Lunch for EWS

Attachment 199500 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/145071

New failing tests:
canvas/philip/tests/initial.reset.2dstate.html
http/tests/security/xss-DENIED-iframe-src-alias.html
fast/canvas/canvas-resize-reset-pixelRatio.html
html5lib/generated/run-tests18-data.html
canvas/philip/tests/initial.reset.same.html
html5lib/generated/run-tests20-data.html
media/media-ended.html
fast/canvas/webgl/context-release-upon-reload.html
fast/frames/frame-set-same-src.html
fast/dom/MutationObserver/observe-attributes.html
http/tests/misc/image-blocked-src-no-change.html
canvas/philip/tests/initial.reset.transform.html
storage/domstorage/events/basic-body-attribute.html
http/tests/navigation/multiple-back-forward-entries.html
fast/forms/number/number-size.html
html5lib/generated/run-scriptdata01-data.html
canvas/philip/tests/initial.reset.path.html
html5lib/generated/run-tests19-data.html
html5lib/generated/run-tests21-data.html
canvas/philip/tests/initial.reset.clip.html
fast/frames/iframe-set-same-src.html
storage/domstorage/events/basic-setattribute.html

[Build Bot]

Created attachment 199508 [details]
Archive of layout-test-results from webkit-ews-01 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-01  Port: mac-mountainlion  Platform: Mac OS X 10.8.2

[Build Bot]

Comment on attachment 199500 [details]
Lunch for EWS

Attachment 199500 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/114521

New failing tests:
html5lib/generated/run-tests18-data.html
canvas/philip/tests/initial.reset.same.html
canvas/philip/tests/initial.reset.2dstate.html
html5lib/generated/run-tests20-data.html
fast/frames/iframe-set-same-src.html
fast/frames/frame-set-same-src.html
http/tests/security/xss-DENIED-iframe-src-alias.html
media/media-ended.html
html5lib/generated/run-scriptdata01-data.html
html5lib/generated/run-tests21-data.html
fast/canvas/canvas-resize-reset-pixelRatio.html
storage/domstorage/events/basic-body-attribute.html
canvas/philip/tests/initial.reset.clip.html
fast/canvas/webgl/context-release-upon-reload.html
fast/forms/number/number-size.html
canvas/philip/tests/initial.reset.path.html
fast/dom/MutationObserver/observe-attributes.html
html5lib/generated/run-tests19-data.html
canvas/philip/tests/initial.reset.transform.html
storage/domstorage/events/basic-setattribute.html

[Build Bot]

Created attachment 199615 [details]
Archive of layout-test-results from webkit-ews-09 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-09  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.2

[Ahmad Saleem]

I don't know whether it is applicable in changed Webkit code, but I don't see change in following place (if I am looking at right place):

https://github.com/WebKit/WebKit/blob/7626eadf0a70b425cf472ec22d17bcb66e16468a/Source/WebCore/dom/Element.cpp#L1900

Just wanted to update on what I understand, if this is not applicable anymore, please mark this as "RESOLVED INVALID" or "RESOLVED WONTFIX". Thanks!

[Ryosuke Niwa]

I'm sure we can implement more perf optimizations but keeping this 9 year old bug won't do us any good.