Issue metadata
Sign in to add a comment
|
Cannot assign to window object when traversing history.
Reported by
storm.pa...@gmail.com,
Feb 9 2017
|
||||||||||||||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36 Steps to reproduce the problem: 1. Create blank page with a embedded script tag that assigns something like window.ID = 42; 2. Open Chrome and observe window.ID === 42 // true 3. Navigate to another page (eg. google.com) and then click the back button 4. Now Chrome throws a Uncaught TypeError: no access What is the expected behavior? Expect that window.ID can be assigned after a page navigation What went wrong? Chrome throws a TypError (Uncaught TypeError: no access) Did this work before? N/A Chrome version: 56.0.2924.87 Channel: stable OS Version: OS X 10.12.3 Flash Version: Shockwave Flash 24.0 r0 Note! This issue is not present when the value is assigned to the global document object. Seems to happen only when assigning to the window object.
,
Feb 10 2017
,
Feb 15 2017
dcheng, is there a good label to put on bugs which are triggered by navigation-type stuff like this? I'm not confident this is just bindings.
,
Feb 15 2017
I can't repro; do you mind providing the test file you're using? My test file is currently: <script>window.ID = 42</script> When I navigate somewhere else and then back, the window.ID assignment doesn't seem to fail.
,
Feb 15 2017
FWIW I can't reproduce this either on Linux 56.0.2924.87 (Official Build) (64-bit) with a page like this: data:text/html,<script>window.ID=42;alert(window.ID === 42);</script><a href="http://google.com">go</a> Could you provide a specific web page we can use to try to reproduce this?
,
Feb 15 2017
Ok, this issue was not as easy to repro as i initially thought. It seems to be related to somekind of coors / session. But as for now i could not yet pinpoint the exact issue. Ill investigate further and will report back here.
,
Feb 22 2017
Thank you for providing more feedback. Adding requester "dominicc@chromium.org" for another review and adding "Needs-Review" label for tracking. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Feb 27 2017
storm.patrik@ could you please provide any latest update regarding this issue. Thank You...
,
Mar 24 2017
I can also replicate this issue in Chrome 56 and 57 with cookie cache all disabled. As said, the issue is very hard to replicate and only seems to occur in an edge case. Here is my steps to replicate: 1. Create a chrome extension that takes over chrome new tab. 2. In the extension new tab, embed an iframe that loads a webpage, for example: http://www.test.com/index.html. 3. In http://www.test.com/index.html, put some JavaScript to access the window variable. For example: setInterval(function(){window.test = 'test'; console.log(window.test)}, 50); 4. Install such extension on your Chrome, so that whenever you open a new tab, you will be shown a page with an iframe that loads http://www.test.com/index.html 5. Open a new tab, at the same time, load http://www.test.com/index.html (the url that is set in the iframe) in the address bar on the same tab. You will have a chance to get the error "Uncaught TypeError: no access" at the lines where it tries to access the window variable (window.test = 'test' or console.log(window.test)) It looks like there is a lock on access to the "window" variable so that when the page in the new tab iframe gets the access to the "window" variable, it locks out the other page with the same domain that you are trying to navigate to on the same tab, in which case you will see "Uncaught TypeError: no access" error.
,
Mar 24 2017
Ah, I didn't know extensions were involved. This is a bug that's fixed (with the fix merged to M58).
,
Mar 31 2017
kkaluri@ Well, i actually just worked around this issue as I really never got to pinpoint the factor involved. Still my browser HAS some extensions, i did not consider them as a culprit... As previously noted this very hard to reproduce, but i still get the bug when accessing my application that does use HTTPS + cookies as a auth mechanism. |
|||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||
Comment 1 by nyerramilli@chromium.org
, Feb 9 2017