Revised test to match behaviour in spec, as revised in issue #245

[faceless2]

These tests originated at https://bugzilla.mozilla.org/show_bug.cgi?id=1624532

I can see they were to fix a regression in FF, but the tested behaviour doesn't match the specification:

• Masks that are missing or invaild are treated as transparent black (ie. nothing is drawn)
• Masks that are display:none are treated as if no mask was specified (note this is not as specified, but matches the resolution in [css-masking] Disabling masks and clipPaths when display is none w3c/fxtf-drafts#245)

Also revised so no red is visible if the tests pass.

@tnikkel

[svgeesus]

Yes, this looks correct to to me per spec (and CSS/SVG WG resolution)

Firefox and Chromium both fail this test.