880592 - Wrong tab count after killed by OS and restored

Status: RESOLVED FIXED

(Firefox for Android Graveyard :: General, defect)

Description

[Brian Nicholson (:bnicholson)]

STR:
1) Open two tabs
2) Select the first tab
3) Wait at least 10 seconds to ensure the session is written to disk
4) Press home
5) Cause Fennec to be killed in the background (either open lots of other apps or run oom-fennec)
6) Reopen fennec

After Fennec is reopened and the session restore, the tab count shows 1 even though two tabs are open.

Comment 1

[Mark Capella [:capella]]

Attachment: Patch (v1)

My bad ... regression caused by bug 874036

Comment 2

[Mark Capella [:capella]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/6ec3af901c39

Comment 3

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/6ec3af901c39

Comment 4

[Mark Capella [:capella]]

Comment on attachment 760180 [details] [diff] [review]
Patch (v1)

[Approval Request Comment]
Bug caused by (feature/regressing bug #): regression re: bug 874036
User impact if declined: Brief incorrect information displayed on restarts after OOM situations.
Testing completed (on m-c, etc.): Yes, STR as attached
Risk to taking this patch (and alternatives if risky): 
String or IDL/UUID changes made by this patch: none

Comment 5

[Alex Keybl [:akeybl]]

(In reply to Mark Capella [:capella] from comment #4)
> Risk to taking this patch (and alternatives if risky): 

Risk evaluation, please :)

Comment 6

[Mark Capella [:capella]]

Sorry, not sure how I missed bugzilla mail with comment 5 ..

This and bug 874036 and bug 868222 (three as a package) are what I'd consider low risk in that they affect UI / displayed information only ... but, they prevent us (temporarily) displaying noticeably wrong information to users which of course is a good thing.

Changes are small and easily backed out, or more likely patched and moved forward in the event of further unseen regressions, though after baking for this long, I can't see that as a serious possibility.

The alternatives are limited, basically go without the fixes and expose wrong information to user scrutiny, or find another way of fixing the underlying problems in the current available time-frame.

Comment 7

[Lukas Blakk [:lsblakk] use ?needinfo]

Comment on attachment 760180 [details] [diff] [review]
Patch (v1)

Thanks Mark, go ahead with uplift.

Comment 8

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-beta/rev/4be587fed6b3