Ian Nartowicz Reporter
	Description • 9 years ago

User Agent: Mozilla/5.0 (X11; Linux i686; rv:38.0) Gecko/20100101 Firefox/38.0
Build ID: 20150514102509

Steps to reproduce:

1. Open an e10s page and set a local storage item (http://www.quirksmode.org/html5/tests/storage.html is handy for this).
2. Open the same page in a non-e10s tab and see if the item is set.
3. Use chrome to look at local storage for the same origin (using domStorageManager.getLocalStorageForPrincipal).


Actual results:

1. Works.
2. The item in the e10s local storage is not visible in the non-e10s page.
3. The item in e10s content local storage is not visible in the chrome localStorage for the same origin.


Expected results:

All local storage areas for the same origin should show the same contents.

	Blake Kaplan (:mrbkap) (inactive)
	Comment 1 • 9 years ago

I can reproduce this.

Honza, I'm looking through dom/storage, and I see the code that sends updates from the child to the parent, but I'm not seeing any sort of code to invalidate caches in either the parent or the child when the database is updated. I'm sure that I'm missing something, but a pointer in the right direction would be much appreciated.

	Honza Bambas (:mayhemer)
	Comment 2 • 9 years ago

(In reply to Blake Kaplan (:mrbkap) (please use needinfo!) from comment #1)
> I can reproduce this.
> 
> Honza, I'm looking through dom/storage, and I see the code that sends
> updates from the child to the parent, but I'm not seeing any sort of code to
> invalidate caches in either the parent or the child when the database is
> updated. I'm sure that I'm missing something, but a pointer in the right
> direction would be much appreciated.

You are not missing anything.  DOMStorage code as is now is designed for just a single child process.  There is no carrying of data updates between multiple child processes running simultaneously and both accessing same origin localStorage at the same time.  

No immediate plan to implement this.

	Blake Kaplan (:mrbkap) (inactive)
	Comment 3 • 9 years ago

Based on comment 2, I'm bumping this back into triage. We'll have to fix this before shipping processCount > 1.