1294386 - [e10s-multi] session storage related test failures with multiple content processes

Status: RESOLVED WORKSFORME

(Firefox :: Session Restore, defect)

Description

[Gabor Krizsanits (INACTIVE)]

This is no surprise we need to fix session storage for e10s-multi. Just want to track the related failing tests here:

browser/components/sessionstore/test/browser_350525.js
browser/components/sessionstore/test/browser_354894_perwindowpb.js
browser/components/sessionstore/test/browser_revive_crashed_bg_tabs.js
browser/components/sessionstore/test/browser_scrollPositions.js

Comment 1

[Ben Kelly [:bkelly, not reviewing]]

If work is being done on session store, you might want to coordinate with Eric.  I think he's been trying to figure out how to fix it to work with bug 1276366 which would solve a lot of our memory leaks.  If we can design the multi-e10s support in a way that also helps there, it would be a huge win.

Comment 2

[Gabor Krizsanits (INACTIVE)]

(In reply to Ben Kelly [:bkelly] from comment #1)
> If work is being done on session store, you might want to coordinate with
> Eric.  I think he's been trying to figure out how to fix it to work with bug
> 1276366 which would solve a lot of our memory leaks.  If we can design the
> multi-e10s support in a way that also helps there, it would be a huge win.

My understanding is that Jan is working on a DOMStorage refactoring that will enable us to let most of the JS implemented session storage to retire. But that might take a while... But thanks, if we end up doing a quick hack in the current session storage to make it e10s-multi compatible as a temporary workaround, I will keep an eye open for bug 1276366.

Comment 3

[Andrew McCreight [:mccr8]]

Looks like these tests were enabled with e10s at some point.