Hi Team, While going through Blink's commits, I came across where we fail over/underflowing Layout Test on WebKit ToT: Blink Commit: https://chromium.googlesource.com/chromium/src.git/+/e18ac62b00618f26ef125554dd8241b2aed711e9 WebKit ToT Result: FAIL large reset does not overflow assert_array_equals: expected property 4 to be 2000000000 but got -1794967296 (expected array [500000000, 1000000000, 1500000000, 2000000000, 2000000000] got [500000000, 1000000000, 1500000000, 2000000000, -1794967296]) FAIL small reset does not underflow assert_array_equals: expected property 4 to be -2000000000 but got 1794967296 (expected array [-500000000, -1000000000, -1500000000, -2000000000, -2000000000] got [-500000000, -1000000000, -1500000000, -2000000000, 1794967296]) Just wanted to raise this bug so we can fix it. Since Vitor and Tim were doing counter-* related fix, so tagging them. Thanks!
<rdar://problem/111743827>
I'll have a look. Thanks for reporting it.
We need to clarify if it should be clamped to the max/min value of the range or if increment should be ignored in this cases. FF and Chrome do the latter, but the spec seems to suggest the former.
https://github.com/w3c/csswg-drafts/issues/9029
As Elika suggested, I think we can aim for compatibility here. I.E: making WebKit also ignore the counter-increment if it would overflow, but making counter-reset clamp to the valid range if it would overflow.
Pull request: https://github.com/WebKit/WebKit/pull/15549
Committed 266817@main (a3717f099f77): <https://commits.webkit.org/266817@main> Reviewed commits have been landed. Closing PR #15549 and removing active labels.