Closed Bug 1644783 Opened 5 years ago Closed 5 years ago

Address bar and page info icons are intermittently wrong

Categories

(Firefox :: Site Identity, defect, P1)

defect

Tracking

()

RESOLVED FIXED
Firefox 79
Tracking Status
firefox-esr68 --- unaffected
firefox-esr78 --- unaffected
firefox77 --- unaffected
firefox78 --- unaffected
firefox79 + fixed

People

(Reporter: emilio, Assigned: ckerschb)

References

(Regression)

Details

(Keywords: regression)

Attachments

(1 file)

I think this has to do with session restore, but I have seen this a couple of times this week.

See the screenshot for an example, where Firefox claims that lando, an https page, is stored locally on my computer.

I hope I'm getting the right component.

Component: Address Bar → Site Identity

This is a possible regression of bug 1570678, but I'm not sure if it really is, since that bug mostly changed cosmetics and not the data we use to display that information.

Would be nice to get at least halfway reliable STR for this :/

Reported on reddit as well.
Reproduces using the URL: https://www.dropbox.com/smart-workspace

6:27.54 INFO: Last good revision: 03c4c517ee37653ee58a04b9feb5947998f123b5
6:27.54 INFO: First bad revision: d9608e8bff0c0eed2ca6399741cafd40a56c1adb
6:27.54 INFO: Pushlog:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=03c4c517ee37653ee58a04b9feb5947998f123b5&tochange=d9608e8bff0c0eed2ca6399741cafd40a56c1adb

Has Regression Range: --- → yes

[Tracking Requested - why for this release]:
Primary security UI broken

Thanks!

(cc Chris since needinfo is blocked)

Chris, can you take a look at this? We use isSecure (and isPotentiallyTrustworthy) in the front-end to distinguish secure connections and potentially trustworthy local resources. I'm really not sure why this issue happens so intermittently/rarely though, it feels like your change should have made it happen more often. I guess that's also why no tests for the UI failed. It might be good to figure out what makes https://www.dropbox.com/smart-workspace different and write a test for it...

In any case, we should solve this for 79, we can't ship like that. Backing out bug 1633338 seems viable.

Severity: -- → S2
Priority: -- → P1

I'm unable to produce this in Firefox v78, but in Firefox v79 (13-06-2020) https://snipboard.io/1YqU8o.jpg

Found out by accident that this seems to happen (only?) on sites where CSP: sandbox headers are set.

Needinfo Chris :)

Flags: needinfo?(ckerschb)

(In reply to Johann Hofmann [:johannh] from comment #7)

Found out by accident that this seems to happen (only?) on sites where CSP: sandbox headers are set.

Needinfo Chris :)

Ah, thanks for letting me know. That is probably because we use a NullPrincipal for sandboxed loads which apparently causes a problem. I guess it's best if we back the change out and re-consider if we want it at all.

Flags: needinfo?(ckerschb)

Emilio, when you get a chance could you verify the backout of Bug 1633338 has fixed the problem? If so, I guess we could close this bug - thanks!

Flags: needinfo?(emilio)

Yeah, seem to be working, or at least I haven't seen this on the latest build.

Status: NEW → RESOLVED
Closed: 5 years ago
Flags: needinfo?(emilio)
Resolution: --- → WORKSFORME
Assignee: nobody → ckerschb
Resolution: WORKSFORME → FIXED
Target Milestone: --- → Firefox 79
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: