[css-flexbox] Don't include scrollbar extents when computing sizes for percentage resolution
Created attachment 403075 [details] Patch
Please, make sure the patch is not the cause of some of the flakiness detected by the mac2 bots: fast/scrolling/mac/scroll-snapping-in-progress.html
Comment on attachment 403075 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=403075&action=review BTW, about the confusion regarding storing something different than what its name says, in Chromium the solution for that was https://chromium-review.googlesource.com/c/chromium/src/+/1024171/, maybe we should try to port that or something similar into WebKit. > LayoutTests/ChangeLog:8 > + * TestExpectations: Unskipped percentage-heights-004.html. The test is only for heights, wouldn't we need a test for widths too? You're changing both things in the code in I'm not getting it wrong.
(In reply to Javier Fernandez from comment #2) > Please, make sure the patch is not the cause of some of the flakiness > detected by the mac2 bots: > > fast/scrolling/mac/scroll-snapping-in-progress.html That's failing without the patch as well.
(In reply to Manuel Rego Casasnovas from comment #3) > Comment on attachment 403075 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=403075&action=review > > BTW, about the confusion regarding storing something different than what its > name says, in Chromium the solution for that was > https://chromium-review.googlesource.com/c/chromium/src/+/1024171/, maybe we > should try to port that or something similar into WebKit. > > > LayoutTests/ChangeLog:8 > > + * TestExpectations: Unskipped percentage-heights-004.html. > > The test is only for heights, wouldn't we need a test for widths too? You're > changing both things in the code in I'm not getting it wrong. Now that you mention it, the width paths are not reachable. I'll post a prequel patch to remove them.
Committed r263794: <https://trac.webkit.org/changeset/263794>
<rdar://problem/64980642>
(In reply to Manuel Rego Casasnovas from comment #3) > Comment on attachment 403075 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=403075&action=review > > BTW, about the confusion regarding storing something different than what its > name says, in Chromium the solution for that was > https://chromium-review.googlesource.com/c/chromium/src/+/1024171/, maybe we > should try to port that or something similar into WebKit. > > > LayoutTests/ChangeLog:8 > > + * TestExpectations: Unskipped percentage-heights-004.html. > > The test is only for heights, wouldn't we need a test for widths too? You're > changing both things in the code in I'm not getting it wrong. After landing bug 213809 there is no need to test widths as those code paths were not reachable and thus they're removed.