Bug 199632 – Remove support for beforeload on link=prefetch

• Component: Page Loading
• Status: RESOLVED FIXED

[Rob Buis]

Remove support for beforeload on link=prefetch.

[Rob Buis]

Created attachment 373742 [details]
Patch

[Rob Buis]

Created attachment 373746 [details]
Patch

[EWS Watchlist]

Comment on attachment 373746 [details]
Patch

Attachment 373746 [details] did not pass mac-ews (mac):
Output: https://webkit-queues.webkit.org/results/12700667

New failing tests:
fast/dom/HTMLLinkElement/prefetch-beforeload.html

[EWS Watchlist]

Created attachment 373753 [details]
Archive of layout-test-results from ews103 for mac-highsierra

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: ews103  Port: mac-highsierra  Platform: Mac OS X 10.13.6

[EWS Watchlist]

Comment on attachment 373746 [details]
Patch

Attachment 373746 [details] did not pass mac-debug-ews (mac):
Output: https://webkit-queues.webkit.org/results/12700823

New failing tests:
fast/dom/HTMLLinkElement/prefetch-beforeload.html

[EWS Watchlist]

Created attachment 373768 [details]
Archive of layout-test-results from ews114 for mac-highsierra

The attached test failures were seen while running run-webkit-tests on the mac-debug-ews.
Bot: ews114  Port: mac-highsierra  Platform: Mac OS X 10.13.6

[EWS Watchlist]

Comment on attachment 373746 [details]
Patch

Attachment 373746 [details] did not pass win-ews (win):
Output: https://webkit-queues.webkit.org/results/12700922

New failing tests:
fast/dom/HTMLLinkElement/prefetch-beforeload.html

[EWS Watchlist]

Created attachment 373769 [details]
Archive of layout-test-results from ews213 for win-future

The attached test failures were seen while running run-webkit-tests on the win-ews.
Bot: ews213  Port: win-future  Platform: CYGWIN_NT-10.0-17763-3.0.5-338.x86_64-x86_64-64bit

[Ryosuke Niwa]

Per recent discussion, I don't think we want to support either load, beforeload, or error event handlers on prefetch link element. That would create arbitrary length tracking vector.

[Rob Buis]

(In reply to Ryosuke Niwa from comment #9)
> Per recent discussion, I don't think we want to support either load,
> beforeload, or error event handlers on prefetch link element. That would
> create arbitrary length tracking vector.

Right, so this bug seems aligned with that, removing support for beforeload events on link prefetch. I just have a hard time testing beforeload does not send an event, especially if load/error events are not supported either, i.e. what is the stop condition and/or how long do we wait?

[Rob Buis]

Reopening to attach new patch.

[Rob Buis]

Created attachment 373907 [details]
Patch

[EWS Watchlist]

Comment on attachment 373907 [details]
Patch

Attachment 373907 [details] did not pass mac-ews (mac):
Output: https://webkit-queues.webkit.org/results/12714121

New failing tests:
http/wpt/prefetch/beforeload-dynamic.html

[EWS Watchlist]

Created attachment 373909 [details]
Archive of layout-test-results from ews101 for mac-highsierra

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: ews101  Port: mac-highsierra  Platform: Mac OS X 10.13.6

[EWS Watchlist]

Comment on attachment 373907 [details]
Patch

Attachment 373907 [details] did not pass mac-debug-ews (mac):
Output: https://webkit-queues.webkit.org/results/12714134

New failing tests:
http/wpt/prefetch/beforeload-dynamic.html

[EWS Watchlist]

Created attachment 373910 [details]
Archive of layout-test-results from ews114 for mac-highsierra

The attached test failures were seen while running run-webkit-tests on the mac-debug-ews.
Bot: ews114  Port: mac-highsierra  Platform: Mac OS X 10.13.6

[Rob Buis]

Created attachment 373911 [details]
Patch

[youenn fablet]

Comment on attachment 373911 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=373911&action=review

> Source/WebCore/ChangeLog:13
> +        (WebCore::LinkLoader::loadLink):

It is not clear these changes remove support for beforeload for prefetch.
I understand it that we remove the call to shouldLoadLink for prefetch which thus removes the call to before load.
Can you give some more description in the change log?

> Source/WebCore/loader/LinkLoader.cpp:286
> +    if (!params.href.isValid() || !document.frame())

It is initially intriguing to not call shouldLoadLink for prefetch.
It might be clearer to call m_client.shouldLoadLink() in every case in LinkLoader::loadLink and return true early if the link is prefetch.

> Source/WebCore/loader/LinkLoader.cpp:332
> +        return true;

It seems loadLink is always returning true.
Maybe loadLink should be void.

> LayoutTests/http/wpt/prefetch/beforeload-dynamic.html:6
> +</script>

Can we create it in the script element below?

> LayoutTests/http/wpt/prefetch/beforeload.html:5
> +    var t = async_test('Makes sure that prefetch does not call beforeload.');

Ditto.

[Rob Buis]

Created attachment 374008 [details]
Patch

[Rob Buis]

Comment on attachment 373911 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=373911&action=review

>> Source/WebCore/ChangeLog:13
>> +        (WebCore::LinkLoader::loadLink):
> 
> It is not clear these changes remove support for beforeload for prefetch.
> I understand it that we remove the call to shouldLoadLink for prefetch which thus removes the call to before load.
> Can you give some more description in the change log?

Sure! Done.

>> Source/WebCore/loader/LinkLoader.cpp:286
>> +    if (!params.href.isValid() || !document.frame())
> 
> It is initially intriguing to not call shouldLoadLink for prefetch.
> It might be clearer to call m_client.shouldLoadLink() in every case in LinkLoader::loadLink and return true early if the link is prefetch.

Done.

>> Source/WebCore/loader/LinkLoader.cpp:332
>> +        return true;
> 
> It seems loadLink is always returning true.
> Maybe loadLink should be void.

Well spotted, done.

>> LayoutTests/http/wpt/prefetch/beforeload-dynamic.html:6
>> +</script>
> 
> Can we create it in the script element below?

Done.

>> LayoutTests/http/wpt/prefetch/beforeload.html:5
>> +    var t = async_test('Makes sure that prefetch does not call beforeload.');
> 
> Ditto.

Done.

[youenn fablet]

Comment on attachment 374008 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=374008&action=review

> Source/WebCore/html/HTMLLinkElement.cpp:208
> +bool HTMLLinkElement::shouldLoadLink(const LinkLoadParameters& params)

Sorry for my suggestion on the previous patch to do that.
Your initial version seems better.
r=me with shouldLoadLink not taking params.

[Rob Buis]

Created attachment 374199 [details]
Patch

[Rob Buis]

(In reply to youenn fablet from comment #21)
> Comment on attachment 374008 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=374008&action=review
> 
> > Source/WebCore/html/HTMLLinkElement.cpp:208
> > +bool HTMLLinkElement::shouldLoadLink(const LinkLoadParameters& params)
> 
> Sorry for my suggestion on the previous patch to do that.
> Your initial version seems better.
> r=me with shouldLoadLink not taking params.

No problem, done.

[WebKit Commit Bot]

Comment on attachment 374199 [details]
Patch

Clearing flags on attachment: 374199

Committed r247481: <https://trac.webkit.org/changeset/247481>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Radar WebKit Bug Importer]

<rdar://problem/53154857>