See https://codepen.io/thiagoboa/pen/abNxRyE?editors=1100 There shouldn't be a line wrap. (lldb) p m_availableWidth (float) $18 = 104.5625 (lldb) p m_committedWidth + m_uncommittedWidth (float) $17 = 104.562508
Alan: Would LFC fix this?
<rdar://problem/69801790>
(In reply to Myles C. Maxfield from comment #1) > Alan: Would LFC fix this? In LFC the preferred width computation and the actual layout share the same codepath with different constrains, so I'd say yes.
Created attachment 410411 [details] Test reduction
Created attachment 410412 [details] LFC screenshot It is actually fixed in LFC (see screenshot).
Created attachment 410414 [details] Patch
will try to submit a test case. it needs a custom font with special geometries (to trigger the floating point issue)
Comment on attachment 410414 [details] Patch No test?
Committed r267923: <https://trac.webkit.org/changeset/267923> All reviewed patches have been landed. Closing bug and clearing flags on attachment 410414 [details].