I've propagated quirks mode information on some places in Bug 1360488. It was a start but it isn't enough for all quirks mode. Also it didn't make all tests pass. I'll continue to work on this in this bug.

Comment on attachment 8867542 [details] Bug 1364746 - stylo: Continue to propagate quirks mode information to Servo https://reviewboard.mozilla.org/r/139090/#review142318 There ere still some static `NoQuirks` in glue.rs file. But I don't think we need to pass real quirks mode information to them. ::: servo/ports/geckolib/glue.rs:1856 (Diff revision 1) > #[no_mangle] > -pub extern "C" fn Servo_CSSSupports2(property: *const nsACString, value: *const nsACString) -> bool { > +pub extern "C" fn Servo_CSSSupports2(property: *const nsACString, > + value: *const nsACString) -> bool { > let id = get_property_id_from_property!(property, false); > > - parse_property(id, value, unsafe { DUMMY_URL_DATA }, structs::LengthParsingMode::Default).is_ok() > + // FIXME: Other parameters are dummy. I guess it's ok to pass NoQuirks here. In here, other parameters are dummy so I passed just NoQuirks here. Do we need to get actual quirk info for that function?

Finally unexpected passes! https://treeherder.mozilla.org/#/jobs?repo=try&revision=60d641348880d71ede0e4254622fff61a6e625ac

Comment on attachment 8867542 [details] Bug 1364746 - stylo: Continue to propagate quirks mode information to Servo https://reviewboard.mozilla.org/r/139092/#review142346

Servo side: https://github.com/servo/servo/pull/16882

Comment on attachment 8867907 [details] Bug 1364746 - stylo: Update test expectations for unitless length quirk https://reviewboard.mozilla.org/r/139438/#review142988

Pushed by canaltinova@gmail.com: https://hg.mozilla.org/integration/autoland/rev/dad46e25b2d7 stylo: Continue to propagate quirks mode information to Servo r=bholley https://hg.mozilla.org/integration/autoland/rev/3dda5445e08d stylo: Update test expectations for unitless length quirk r=bholley

Pushed by ecoal95@gmail.com: https://hg.mozilla.org/integration/autoland/rev/2a534e65b84f More reftest expectation updates. r=me

Hmm, so I was updating the test expectations without noticing the gecko side of your patch hadn't landed yet. Nazim, if those tests I updated happen to not pass, could you please file a bug to investigate those? They're definitely related to parsing without units, so if they're not passing there's a bug servo-side.

Sure, I'll look into it.

Pushed by ecoal95@gmail.com: https://hg.mozilla.org/integration/autoland/rev/0be21772f2f9 Revert "Bug 1364746: More reftest expectation updates. r=me"

I think this has been fixed.