Bug 87355 – canvas/philip/tests/2d.text.draw.fontface.notinpage.html is flaky/failing because of custom font loading

• Component: Tools / Tests
• Status: ASSIGNED

[Chris Dumez]

canvas/philip/tests/2d.text.draw.fontface.notinpage.html is failing on EFL port. The problems seems be caused the loading of a custom font, which delays the drawing of the text on the canvas.
As a consequence, the text is not drawn yet when the test is checking points color on the canvas.

[Zan Dobersek]

The test now also fails on the GTK platform after the memory cache started to be cleared in between test runs in r126846:
http://trac.webkit.org/changeset/126846

[Alexey Proskuryakov]

This test is currently flaky on Mac, both WK1 and WK2.

[Chris Dumez]

Bug 96835 is related.

[Chris Dumez]

I'm working on a patch proposal for this. I'll hopefully upload it soon, sorry for the delay.

[Chris Dumez]

Created attachment 176056 [details]
Patch

Here is a patch proposal. I confirmed that it makes LayoutTests/canvas/philip/tests/2d.text.draw.fontface.notinpage.html pass consistently and does not seem to cause any regression.

I open to suggestions if someone has a better idea to solve this.

[Peter Beverloo (cr-android ews)]

Comment on attachment 176056 [details]
Patch

Attachment 176056 [details] did not pass cr-android-ews (chromium-android):
Output: http://queues.webkit.org/results/14982796

[WebKit Review Bot]

Comment on attachment 176056 [details]
Patch

Attachment 176056 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/14980928

[Tony Chang]

Does the canvas spec say what should happen when trying to draw text that hasn't loaded yet?  Waiting for the font sounds wrong.

[Chris Dumez]

I took a closer look at the latest draft (http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#2dcontext):

"If a font is used before it is fully loaded, or if the font style source object does not have that font in scope at the time the font is to be used, then it must be treated as if it was an unknown font, falling back to another as described by the relevant CSS specifications. [CSSFONTS]"

So, it seems that you are right. We apparently shouldn't wait for the font but instead fallback to another. I'm not sure this is going to help with canvas/philip/tests/2d.text.draw.fontface.notinpage.html though as it may rely on the custom font to pass (I'll need to try).

[Darin Adler]

(In reply to comment #9)
> We apparently shouldn't wait for the font but instead fallback to another. I'm not sure this is going to help with canvas/philip/tests/2d.text.draw.fontface.notinpage.html though as it may rely on the custom font to pass (I'll need to try).

If so, then it seems the test is flawed and requires behavior that’s not part of the spec.

[Chris Dumez]

(In reply to comment #10)
> (In reply to comment #9)
> > We apparently shouldn't wait for the font but instead fallback to another. I'm not sure this is going to help with canvas/philip/tests/2d.text.draw.fontface.notinpage.html though as it may rely on the custom font to pass (I'll need to try).
> 
> If so, then it seems the test is flawed and requires behavior that’s not part of the spec.

Yes, I agree.