839399 - [Captive Portal] JS Error if the network error happen while fetching captive portal detection web page

Status: RESOLVED WORKSFORME

(Firefox OS Graveyard :: General, defect)

Description

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

The following JS error shows in logcat with default captive portal detection site point to http://localhost/test.html
E/GeckoConsole(  771): [JavaScript Error: "self._runningRequest is null" {file: "jar:file:///system/b2g/omni.ja!/components/captivedetect.js" line: 301}]

Root Cause:
XMLHttpRequest.onreadystatechange() will still be invoked with readyState == DONE and status == 0 before onerror(), therefore, the retry mechanism will have this error since related attributes are cleaned up by onerror().

Comment 1

[Jason Smith [:jsmith]]

Is this worth noming for shira? What's the user impact here if this isn't fixed?

Comment 2

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

(In reply to Jason Smith [:jsmith] from comment #1)
> Is this worth noming for shira? What's the user impact here if this isn't
> fixed?
The retry mechanism for captive portal detection will not take effect under a sophisticated test case. User is not impacted by this JS error in their daily use. However, I already came out a simple patch for resolving this bug and I'd like to ship the captive portal detection in an error-free state.

Comment 3

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

Attachment: Do not invoke onredirectorerror callback while network failure

For network failure on XMLHttpRequest, onreadystatechange will still be invoked with readyState = 4 and status = 0 prior to onerror callback. We should not invoke the onredirectorerror callback and should let onerror callback handle it.

Comment 4

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

Attachment: Test case

xpcshell test case for reproducing this scenario.

Comment 5

[Stephen Donner [:stephend] Not actively reading bugmail]

Attachment: logcat from simple disconnect/reconnect/launch Marketplace scenario

Are we saying most normal users won't see this?  In my case (a simple scenario), I:

1. Disconnected from Mozilla Guest
2. Reconnected
3. Launched Marketplace

In our automation, too, we see:

E/GeckoConsole(  109): [JavaScript Error: "self._runningRequest is null" {file: "jar:file:///system/b2g/omni.ja!/components/captivedetect.js" line: 301}]

quite often - could there be anything else that caused this failure to launch Marketplace?

E/GeckoConsole(  109): Content JS WARN at app://system.gaiamobile.org/js/window.js:40 in anonymous: app of [https://marketplace.firefox.com/telefonica/] got a mozbrowsererror event.

Comment 6

[Stephen Donner [:stephend] Not actively reading bugmail]

(In reply to Stephen Donner [:stephend] from comment #5)

I spun this off as bug 840612.

Comment 7

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

Patches were merged with original modification of captive portal detection in bug 752982. This one can be closed without any further modification.