Simple line layout: Use simplified text measuring when possible.

• Component: Layout and Rendering
• Status: RESOLVED FIXED

[zalan]

Measuring simple text is simple.

[Radar WebKit Bug Importer]

<rdar://problem/30364907>

[zalan]

Created attachment 300644 [details]
Patch

[Jon Lee]

Comment on attachment 300644 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=300644&action=review

> Source/WebCore/ChangeLog:11
> +        Performance test is added.

...and what does it show?

[zalan]

(In reply to comment #3)
> Comment on attachment 300644 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=300644&action=review
> 
> > Source/WebCore/ChangeLog:11
> > +        Performance test is added.
> 
> ...and what does it show?
oh the changelog is not ready. This is only for EWS.
(it shows how long it takes to layout the content in milliseconds)

[Build Bot]

Comment on attachment 300644 [details]
Patch

Attachment 300644 [details] did not pass mac-debug-ews (mac):
Output: http://webkit-queues.webkit.org/results/3005928

New failing tests:
fast/text/font-variant-ligatures.html
fast/text/system-font-features.html
http/tests/misc/acid3.html
css3/font-variant-font-face-override.html

[Build Bot]

Created attachment 300653 [details]
Archive of layout-test-results from ews112 for mac-elcapitan

The attached test failures were seen while running run-webkit-tests on the mac-debug-ews.
Bot: ews112  Port: mac-elcapitan  Platform: Mac OS X 10.11.6

[zalan]

Created attachment 300664 [details]
Patch

[zalan]

Created attachment 300665 [details]
Patch

[zalan]

Created attachment 300678 [details]
Patch

[zalan]

Created attachment 300723 [details]
Patch

[zalan]

Created attachment 300727 [details]
Patch

[Antti Koivisto]

Comment on attachment 300727 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=300727&action=review

> Source/WebCore/platform/graphics/WidthCache.h:133
> +    {
> +        if (MemoryPressureHandler::singleton().isUnderMemoryPressure())
> +            return nullptr;
> +
> +        if (static_cast<unsigned>(text.length()) > SmallStringKey::capacity())
> +            return nullptr;
> +
> +        if (m_countdown > 0) {
> +            --m_countdown;
> +            return nullptr;
> +        }

Could we factor these tests into a function shared with the other add()?

[zalan]

(In reply to comment #12)
> Comment on attachment 300727 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=300727&action=review
> 
> > Source/WebCore/platform/graphics/WidthCache.h:133
> > +    {
> > +        if (MemoryPressureHandler::singleton().isUnderMemoryPressure())
> > +            return nullptr;
> > +
> > +        if (static_cast<unsigned>(text.length()) > SmallStringKey::capacity())
> > +            return nullptr;
> > +
> > +        if (m_countdown > 0) {
> > +            --m_countdown;
> > +            return nullptr;
> > +        }
> 
> Could we factor these tests into a function shared with the other add()?
Only the first 2 checks could be merged and the slow path has the length() check as the last one and I assume the order is fine-tuned.
Let me address this in a follow-up patch.

[WebKit Commit Bot]

Comment on attachment 300727 [details]
Patch

Clearing flags on attachment: 300727

Committed r211738: <http://trac.webkit.org/changeset/211738>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.