238126 2022-03-20 08:41:05 -0700 [css-cascade] No need to defer applying text-decoration properties 2022-03-21 12:39:30 -0700 1 1 1 Unclassified WebKit CSS WebKit Nightly Build Unspecified Unspecified RESOLVED FIXED InRadar P2 Normal --- 237175 238125 1 obrufau obrufau darin koivisto webkit-bug-importer oldest_to_newest 1853065 0 obrufau 2022-03-20 08:41:05 -0700 In Source/WebCore/style/PropertyCascade.cpp, shouldApplyPropertyInParseOrder() returns true for - CSSPropertyWebkitTextDecoration - CSSPropertyTextDecorationLine - CSSPropertyTextDecorationStyle - CSSPropertyTextDecorationColor - CSSPropertyTextDecorationSkip - CSSPropertyTextDecorationSkipInk - CSSPropertyTextUnderlinePosition - CSSPropertyTextUnderlineOffset - CSSPropertyTextDecorationThickness - CSSPropertyTextDecoration This was previously necessary for text-decoration-line and text-decoration, since they were implemented as longhands that shared a computed value. But that's no longer the case, text-decoration became a shorthand in bug 237175. AFAIK -webkit-text-decoration has always been a shorthand since it was implemented in bug 92000, so having it in the list it's pointless, only longhands matter. text-decoration-skip became a shorthand in bug 230244, so it's also pointless. And the other longhands seem unnecessary too, since they don't share a computed style with other properties. 1853066 1 455196 obrufau 2022-03-20 08:54:10 -0700 Created attachment 455196 Patch 1853079 2 obrufau 2022-03-20 10:40:48 -0700 PTAL, trivial patch. 1853224 3 darin 2022-03-21 09:29:56 -0700 Retrying to see if the iOS-wk2 failure goes away. 1853251 4 455196 obrufau 2022-03-21 10:53:29 -0700 Comment on attachment 455196 Patch I already retried iOS-wk2 a few times, but fast/text/text-shadow-ink-overflow-missing.html keeps failing all the time, and not just with my patch: https://ews-build.webkit.org/#/builders/iOS-15-Simulator-WK2-Tests-EWS So it's unrelated, cq+ 1853256 5 ews-feeder 2022-03-21 11:01:18 -0700 Committed r291568 (248669@main): <https://commits.webkit.org/248669@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 455196. 1853258 6 webkit-bug-importer 2022-03-21 11:02:20 -0700 <rdar://problem/90578051> 1853259 7 darin 2022-03-21 11:02:58 -0700 That's not how EWS is supposed to work. If it fails both with and without the patch it's not supposed to be red. Look at an example like this one: https://ews-build.webkit.org/#/builders/68/builds/11070 Note that it says the test *passes* without the patch it was evaluating. I wonder why EWS comes to that conclusion. 1853293 8 obrufau 2022-03-21 12:30:20 -0700 (In reply to Darin Adler from comment #7) > That's not how EWS is supposed to work. If it fails both with and without > the patch it's not supposed to be red. Look at an example like this one: > > https://ews-build.webkit.org/#/builders/68/builds/11070 > > Note that it says the test *passes* without the patch it was evaluating. I > wonder why EWS comes to that conclusion. I think it may be because when running the tests without the patch, it only runs the tests that previously failed, not all of them. Maybe this uses less resources on the machine or something, and affects the result? So it's not a matter of whether the patch is applied or not, but the number tests running at the same time. These just happen to be correlated. 1853295 9 darin 2022-03-21 12:39:30 -0700 Seems like we need to change EWS to run the tests that failed *with* the patch too then, otherwise the A/B is invalid. 455196 2022-03-20 08:54:10 -0700 2022-03-21 11:01:20 -0700 Patch bug-238126-20220320165409.patch text/plain 2869 obrufau U3VidmVyc2lvbiBSZXZpc2lvbjogMjkxNTQwCmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9D aGFuZ2VMb2cgYi9Tb3VyY2UvV2ViQ29yZS9DaGFuZ2VMb2cKaW5kZXggODA1ZWYxYzliZmM0MDA3 ZWVhMzA2ZDc2ZjhlZGEzZTQ0YjI2ODkzOC4uMTdhMTdhMmRiZDUyYmM1ZmRjY2U2NmJhMzM0ZTRm ODYyNmUxMTg1MiAxMDA2NDQKLS0tIGEvU291cmNlL1dlYkNvcmUvQ2hhbmdlTG9nCisrKyBiL1Nv dXJjZS9XZWJDb3JlL0NoYW5nZUxvZwpAQCAtMSwzICsxLDQwIEBACisyMDIyLTAzLTIwICBPcmlv bCBCcnVmYXUgIDxvYnJ1ZmF1QGlnYWxpYS5jb20+CisKKyAgICAgICAgW2Nzcy1jYXNjYWRlXSBE b24ndCBkZWZlciBhcHBseWluZyB0ZXh0IGRlY29yYXRpb24gcHJvcGVydGllcworICAgICAgICBo dHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1Zy5jZ2k/aWQ9MjM4MTI2CisKKyAgICAgICAg UmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgc2hvdWxkQXBwbHlQcm9wZXJ0 eUluUGFyc2VPcmRlcigpIHdhcyByZXR1cm5pbmcgdHJ1ZSBmb3IgdGhlc2U6CisgICAgICAgICAt IHdlYmtpdC10ZXh0LWRlY29yYXRpb24KKyAgICAgICAgIC0gdGV4dC1kZWNvcmF0aW9uLWxpbmUK KyAgICAgICAgIC0gdGV4dC1kZWNvcmF0aW9uLXN0eWxlCisgICAgICAgICAtIHRleHQtZGVjb3Jh dGlvbi1jb2xvcgorICAgICAgICAgLSB0ZXh0LWRlY29yYXRpb24tc2tpcAorICAgICAgICAgLSB0 ZXh0LWRlY29yYXRpb24tc2tpcC1pbmsKKyAgICAgICAgIC0gdGV4dC11bmRlcmxpbmUtcG9zaXRp b24KKyAgICAgICAgIC0gdGV4dC11bmRlcmxpbmUtb2Zmc2V0CisgICAgICAgICAtIHRleHQtZGVj b3JhdGlvbi10aGlja25lc3MKKyAgICAgICAgIC0gdGV4dC1kZWNvcmF0aW9uCisKKyAgICAgICAg VGhpcyB3YXMgcHJldmlvdXNseSBuZWVkZWQgZm9yIHRleHQtZGVjb3JhdGlvbi1saW5lIGFuZCB0 ZXh0LWRlY29yYXRpb24sCisgICAgICAgIHNpbmNlIHRoZXkgd2VyZSBpbXBsZW1lbnRlZCBhcyBs b25naGFuZHMgdGhhdCBzaGFyZWQgYSBjb21wdXRlZCB2YWx1ZS4KKyAgICAgICAgQnV0IHRoYXQn cyBubyBsb25nZXIgdGhlIGNhc2UsIHRleHQtZGVjb3JhdGlvbiBiZWNhbWUgYSBzaG9ydGhhbmQg aW4gYnVnCisgICAgICAgIDIzNzE3NS4KKworICAgICAgICBBRkFJSyAtd2Via2l0LXRleHQtZGVj b3JhdGlvbiBoYXMgYWx3YXlzIGJlZW4gYSBzaG9ydGhhbmQgc2luY2UgaXQgd2FzCisgICAgICAg IGltcGxlbWVudGVkIGluIGJ1ZyA5MjAwMCwgc28gaGF2aW5nIGl0IGluIHRoZSBsaXN0IGl0J3Mg cG9pbnRsZXNzLAorICAgICAgICBvbmx5IGxvbmdoYW5kcyBtYXR0ZXIuIEFuZCB0ZXh0LWRlY29y YXRpb24tc2tpcCBiZWNhbWUgYSBzaG9ydGhhbmQgaW4KKyAgICAgICAgYnVnIDIzMDI0NCwgc28g aXQncyBhbHNvIHBvaW50bGVzcy4KKworICAgICAgICBUaGUgb3RoZXIgbG9uZ2hhbmRzIHNlZW0g dW5uZWNlc3NhcnkgdG9vLCBzaW5jZSB0aGV5IGRvbid0IHNoYXJlIGEKKyAgICAgICAgY29tcHV0 ZWQgc3R5bGUgd2l0aCBvdGhlciBwcm9wZXJ0aWVzLgorCisgICAgICAgIE5vIHRlc3Qgc2luY2Ug dGhlcmUgc2hvdWxkIGJlIG5vIG9ic2VydmFibGUgY2hhbmdlIGluIGJlaGF2aW9yLgorCisgICAg ICAgICogc3R5bGUvUHJvcGVydHlDYXNjYWRlLmNwcDoKKyAgICAgICAgKFdlYkNvcmU6OlN0eWxl OjpzaG91bGRBcHBseVByb3BlcnR5SW5QYXJzZU9yZGVyKToKKwogMjAyMi0wMy0yMCAgT3Jpb2wg QnJ1ZmF1ICA8b2JydWZhdUBpZ2FsaWEuY29tPgogCiAgICAgICAgIEZpeCBDU1MgY2FzY2FkZSBy ZWdhcmRpbmcgbG9naWNhbCBwcm9wZXJ0aWVzCmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9z dHlsZS9Qcm9wZXJ0eUNhc2NhZGUuY3BwIGIvU291cmNlL1dlYkNvcmUvc3R5bGUvUHJvcGVydHlD YXNjYWRlLmNwcAppbmRleCA4NTkzZjcwOWRjOTMxM2E4YjM3MDMzNTUwZjNiNTg5Mzc5OTJhMjkx Li4yNGQ4MGIxNzFmZWNiZjU4MGFmNmU5NzVhNmQ5ODE1NGVlZGMxZDg2IDEwMDY0NAotLS0gYS9T b3VyY2UvV2ViQ29yZS9zdHlsZS9Qcm9wZXJ0eUNhc2NhZGUuY3BwCisrKyBiL1NvdXJjZS9XZWJD b3JlL3N0eWxlL1Byb3BlcnR5Q2FzY2FkZS5jcHAKQEAgLTU2LDE2ICs1Niw2IEBAIHN0YXRpYyBp bmxpbmUgYm9vbCBzaG91bGRBcHBseVByb3BlcnR5SW5QYXJzZU9yZGVyKENTU1Byb3BlcnR5SUQg cHJvcGVydHlJRCkKICAgICBjYXNlIENTU1Byb3BlcnR5Qm9yZGVySW1hZ2VXaWR0aDoKICAgICBj YXNlIENTU1Byb3BlcnR5V2Via2l0Qm94U2hhZG93OgogICAgIGNhc2UgQ1NTUHJvcGVydHlCb3hT aGFkb3c6Ci0gICAgY2FzZSBDU1NQcm9wZXJ0eVdlYmtpdFRleHREZWNvcmF0aW9uOgotICAgIGNh c2UgQ1NTUHJvcGVydHlUZXh0RGVjb3JhdGlvbkxpbmU6Ci0gICAgY2FzZSBDU1NQcm9wZXJ0eVRl eHREZWNvcmF0aW9uU3R5bGU6Ci0gICAgY2FzZSBDU1NQcm9wZXJ0eVRleHREZWNvcmF0aW9uQ29s b3I6Ci0gICAgY2FzZSBDU1NQcm9wZXJ0eVRleHREZWNvcmF0aW9uU2tpcDoKLSAgICBjYXNlIENT U1Byb3BlcnR5VGV4dERlY29yYXRpb25Ta2lwSW5rOgotICAgIGNhc2UgQ1NTUHJvcGVydHlUZXh0 VW5kZXJsaW5lUG9zaXRpb246Ci0gICAgY2FzZSBDU1NQcm9wZXJ0eVRleHRVbmRlcmxpbmVPZmZz ZXQ6Ci0gICAgY2FzZSBDU1NQcm9wZXJ0eVRleHREZWNvcmF0aW9uVGhpY2tuZXNzOgotICAgIGNh c2UgQ1NTUHJvcGVydHlUZXh0RGVjb3JhdGlvbjoKICAgICAgICAgcmV0dXJuIHRydWU7CiAgICAg ZGVmYXVsdDoKICAgICAgICAgcmV0dXJuIENTU1Byb3BlcnR5Ojppc0luTG9naWNhbFByb3BlcnR5 R3JvdXAocHJvcGVydHlJRCk7Cg==