Bug 87145 – REGRESSION (r117815) fast/css/zoom-in-length-round-trip.html fails on Mac bots

• Component: CSS
• Status: RESOLVED FIXED

[Brady Eidson]

REGRESSION (r117815) fast/css/zoom-in-length-round-trip.html fails on Mac bots

Link to change:
http://trac.webkit.org/changeset/117815

Link to example of failure:
http://build.webkit.org/results/Lion%20Release%20(Tests)/r117988%20(8548)/results.html

117815 purported to be only about enabling a feature on Chromium, so it seems odd it could have caused this.

But it never failed up to and including 117813, then has been failing started in 117815.

117814 cannot possibly have affected Mac as it only touched a Chromium.cpp file - (http://trac.webkit.org/changeset/117814) - so 117815 must be it.

[Brady Eidson]

Skipped in http://trac.webkit.org/changeset/118003

Please unskip after you have resolved the regression.

[Eric Seidel (no email)]

It's possible, but it would have come from an earlier change in teh subpixel change I think.

[Brady Eidson]

(In reply to comment #2)
> It's possible, but it would have come from an earlier change in teh subpixel change I think.

I'm having a hard time grok'ing this sentence.

All I know is that from 117813 back the test was fine.

[Levi Weintraub]

(In reply to comment #3)
> (In reply to comment #2)
> > It's possible, but it would have come from an earlier change in teh subpixel change I think.
> 
> I'm having a hard time grok'ing this sentence.
> 
> All I know is that from 117813 back the test was fine.

The numbers being returned are correct, it looks like a test issue.

[Emil A Eklund]

Created attachment 143337 [details]
Patch

[Eric Seidel (no email)]

Comment on attachment 143337 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=143337&action=review

> LayoutTests/fast/css/zoom-in-length-round-trip.html:33
> +            if (hasSubpixelSupport && (target.offsetWidth == 1 || target.offsetWidth == 2) && (target.offsetHeight == 10 || target.offsetHeight == 11))

I don'dt quite understnad thsi complicated check.

[Emil A Eklund]

(In reply to comment #6)
> (From update of attachment 143337 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=143337&action=review
> 
> > LayoutTests/fast/css/zoom-in-length-round-trip.html:33
> > +            if (hasSubpixelSupport && (target.offsetWidth == 1 || target.offsetWidth == 2) && (target.offsetHeight == 10 || target.offsetHeight == 11))
> 
> I don'dt quite understnad thsi complicated check.

That's why I added the comment...

[Levi Weintraub]

Comment on attachment 143337 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=143337&action=review

Please also unskip this test on Mac before landing.

> LayoutTests/fast/css/zoom-in-length-round-trip.html:33
> +            if (hasSubpixelSupport && (target.offsetWidth == 1 || target.offsetWidth == 2) && (target.offsetHeight == 10 || target.offsetHeight == 11))

This is ugly to have to do, but okay.

[Brady Eidson]

Even though Levi r+'ed, please unskip the test as well.

[Emil A Eklund]

(In reply to comment #9)
> Even though Levi r+'ed, please unskip the test as well.

Will do!

[Levi Weintraub]

(In reply to comment #8)
> (From update of attachment 143337 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=143337&action=review
> 
> Please also unskip this test on Mac before landing.

^^^ :)

[Emil A Eklund]

Committed r118013: <http://trac.webkit.org/changeset/118013>

[Levi Weintraub]

*** Bug 87099 has been marked as a duplicate of this bug. ***