Bug 89458
| Summary: | [EFL] Enable SUBPIXEL_LAYOUT feature flag for EFL | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Dominik Röttsches (drott) <d-r> |
| Component: | WebKit EFL | Assignee: | Dominik Röttsches (drott) <d-r> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | eae, gyuyoung.kim, laszlo.gombos, leviw, lucas.de.marchi, ostap73, s.choi, sw0524.lee, tmpsantos |
| Priority: | P2 | Keywords: | WebExposed |
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Bug Depends on: | 89572, 90169, 102374, 103121, 103129, 106774, 107548 | ||
| Bug Blocks: | 85532 | ||
Dominik Röttsches (drott)
We'd like to follow suit with bug 85532, enabling sub-pixel layout for the EFL port.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Dominik Röttsches (drott)
First obstacle, in a number of cases
setLogicalTop(roundToInt(top + maxAscent - fontMetrics.ascent(baselineType)));
inside
InlineFlowBox::placeBoxesInBlockDirection
actually results in -1.
maxAscent is 840/60 = 14, while fontMetrics.ascent (and fontMetrics.floatAscent fwiw) return 15.
Will investigate - but if you have any ideas, Levi, Emil - let me know.
Dominik Röttsches (drott)
(In reply to comment #1)
> First obstacle, in a number of cases
> setLogicalTop(roundToInt(top + maxAscent - fontMetrics.ascent(baselineType)));
> inside
> InlineFlowBox::placeBoxesInBlockDirection
> actually results in -1.
To clarify: ...leading to hundreds of test cases failing to show expected results just by this -1 offset - they previously had the logialTop just at 0.
Dominik Röttsches (drott)
Freetype updated and feature enabled in bug 106774, new baselines completely uploaded and the bots are returning back to normal. Closing bug.
Dominik Röttsches (drott)
Big thanks, Thiago, for helping with the rebaselines!
Emil A Eklund
Awesome, thanks for making this happen Dominik!