[Webkit-unassigned] [Bug 199866] New: WebKit-based browser hangs while parsing large JS file

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Jul 17 09:04:26 PDT 2019


            Bug ID: 199866
           Summary: WebKit-based browser hangs while parsing large JS file
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: Macintosh
                OS: macOS 10.14
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: New Bugs
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: redbrogdon at google.com

I've run into an issue attempting to load a large JS file (the compiled-to-JavaScript version of the Dart SDK) in Safari, and I'm unable to replicate the problem in other, non-WebKit browsers.

Loading the file as if it were a single script with no requireJS-style wrapper or ES6 module import works perfectly well (~5MB of JS in <300ms). When I add the `define` wrapper needed for requireJS or add an export statement and attempt to load the file via an ES6 module, however, Safari freezes for ~10 seconds during hard reloads. I'm unable to interact with the page at all. Once the script is loaded, I can usually reload the page without incurring the delay (I'm guessing whatever in-memory representation is made of the script file hangs out in between loads).

I knocked out a GitHub repo to demonstrate the issue at https://github.com/RedBrogdon/parsing_speed. It has four HTML files, all of which try different techniques for loading the file. They're marked for which have a delay in Safari/WebKit and which do not. I've tested that code with Safari v12.1.1 (14607. and the MiniBrowser shipped with build 247494 of WebKit. Both show the same behavior, which I've been unable to replicate with any other browser.

I'm not a JS engine expert by any means, but I'm guessing there's a parsing optimization that's not available when the code in my file is moved out of global scope and into a function/module, and it's possible that the behavior I'm seeing is WebKit working as intended. If that's the case and you can suggest some way for me to modify my code (or even point me to a guide that explains when/how parsing can be optimized), I'd appreciate it.

You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20190717/70e0d03d/attachment-0001.html>

More information about the webkit-unassigned mailing list