RESOLVED INVALID250072
Resize event on iframe fired before rAF on parent document 
https://bugs.webkit.org/show_bug.cgi?id=250072
Summary Resize event on iframe fired before rAF on parent document
Rune Lillesveen
Reported 2023-01-04 03:44:45 PST
In the case below the rAF on the parent document should be handled before the iframe re-layout is done, which means the resize event handler should be run after the parent document rAF handler. Safari outputs the "resize" before the "rAF": <!doctype html> <iframe id="frm" width="100"></iframe> <script> setTimeout(() => { frm.contentWindow.addEventListener('resize', () => console.log("resize")); requestAnimationFrame(() => console.log("rAF")); frm.style.width = '200px'; }, 1000); </script> Found this while looking at failures in Chrome for: http://wpt.live/css/mediaqueries/media-query-matches-in-iframe.html Some of the expectations there seem wrong.
Attachments
Add attachment proposed patch, testcase, etc.
Radar WebKit Bug Importer
Comment 1 2023-01-04 16:55:04 PST
<rdar://problem/103893924>
Simon Fraser (smfr)
Comment 2 2023-01-04 20:08:45 PST
https://html.spec.whatwg.org/multipage/webappapis.html#event-loop-processing-model says that the resize (step 7) happens before the animation frame callbacks (step 13).
Rune Lillesveen
Comment 3 2023-01-09 04:45:26 PST
> https://html.spec.whatwg.org/multipage/webappapis.html#event-loop-processing-model says that the resize (step 7) happens before the animation frame callbacks (step 13). Yes, but the resize of the iframe is a result of the layout of the top document that happens after the first rAF callback. The resize event should then trigger before the next rAF, IIUC?
Ryosuke Niwa
Comment 4 2023-03-22 14:56:28 PDT
(In reply to Rune Lillesveen from comment #3) > > https://html.spec.whatwg.org/multipage/webappapis.html#event-loop-processing-model says that the resize (step 7) happens before the animation frame callbacks (step 13). > > Yes, but the resize of the iframe is a result of the layout of the top > document that happens after the first rAF callback. The resize event should > then trigger before the next rAF, IIUC? I don't see how the spec reads that way. Fundamentally, spec basically assumes that the resize happens immediately once the width is set, not when rAF happens.
Note You need to log in before you can comment on or make changes to this bug.