1079707 - Accessing IPv6 URLs without typing the protocol and the first bracket throws a JS error in the Browser Console

Status: VERIFIED FIXED

(Firefox :: Address Bar, defect)

Description

[Andrei Vaida [:avaida]]

Note: this is a follow up bug for Bug 494092.

Reproducible on:
* Nightly 35.0a1 (2014-09-15), 
* Windows 7 64-bit, Ubuntu 14.04 LTS 32-bit, Mac OS X 10.9.5.

STR:
1. Launch Firefox with a clean profile.
2. Open the Browser Console.
3. Select the Location Bar and type in the following IPv6 address: '[FE80::0202:B3FF:FE1E:8329' (w/o apostrophe).
4. Press the <enter> key and look at the logs displayed in the Browser Console.

ER:
There are no errors or unexpected warnings displayed for that request. A search is performed for that invalid IPv6 URL.

AR:
A search is indeed performed for that invalid IPv6 URL, but the following JS error is displayed in the Browser Console: 
> NS_ERROR_UNKNOWN_HOST: Component returned failure code: 0x804b001e 
> (NS_ERROR_UNKNOWN_HOST) [nsIDNSService.asyncResolve] 
> in browser.js:10658

Additional notes:
- This issue *might be a regression*, but so far I was unable to find a reliable regression window.
- Per Bug 494092 Comment 36, a potential regressor might be Bug 1048618.
- This issue is reproducible across platforms (Windows 7 x64, Ubuntu 14 x86, OS X 10.9.5).

Comment 1

[:Gijs (he/him)]

We should try...catch the asyncResolve call.

Comment 2

[:Gijs (he/him)]

Attachment: asyncresolve can throw,

Comment 3

[Alex Bardas :alexbardas]

Comment on attachment 8504506 [details] [diff] [review]
asyncresolve can throw,

Review of attachment 8504506 [details] [diff] [review]:
-----------------------------------------------------------------

I guess the comment can be a bit improved, but it can be easily understood from the context.

Comment 4

[:Gijs (he/him)]

(In reply to Alex Bardas :alexbardas from comment #3)
> Comment on attachment 8504506 [details] [diff] [review]
> asyncresolve can throw,
> 
> Review of attachment 8504506 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> I guess the comment can be a bit improved, but it can be easily understood
> from the context.

What would you prefer for the comment? :-)

Comment 5

[Alex Bardas :alexbardas]

(In reply to :Gijs Kruitbosch from comment #4)
> What would you prefer for the comment? :-)

I'm not very opinionated here, but I was thinking of something like

"Do nothing if the URL is invalid (we don't want to show a notification in that case)."

to be more consistent with other comments.

Comment 6

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/044923edbc61

Comment 7

[Andrei Vaida [:avaida]]

Nightly 36.0a1 (2014-10-17), Windows 7 64-bit

I'm seeing a different JS error thrown in the Browser Console now, for the exact same steps from Comment 0:
> [Exception... "Component returned failure code: 0x804b000a (NS_ERROR_MALFORMED_URI) 
> [nsIURIFixup.createFixupURI]"  nsresult: "0x804b000a (NS_ERROR_MALFORMED_URI)"  
> location: "JS frame :: resource:///modules/PlacesUIUtils.jsm :: PUIU_createFixedURI :: line 97"
> data: no]
in urlbarbindings.xml:332.

I assume this isn't intended. Gijs, what's your take on this?

Comment 8

[:Gijs (he/him)]

(In reply to Andrei Vaida, QA [:avaida] from comment #7)
> Nightly 36.0a1 (2014-10-17), Windows 7 64-bit
> 
> I'm seeing a different JS error thrown in the Browser Console now, for the
> exact same steps from Comment 0:
> > [Exception... "Component returned failure code: 0x804b000a (NS_ERROR_MALFORMED_URI) 
> > [nsIURIFixup.createFixupURI]"  nsresult: "0x804b000a (NS_ERROR_MALFORMED_URI)"  
> > location: "JS frame :: resource:///modules/PlacesUIUtils.jsm :: PUIU_createFixedURI :: line 97"
> > data: no]
> in urlbarbindings.xml:332.
> 
> I assume this isn't intended. Gijs, what's your take on this?

That's a totally different part of the code to do with the urlbar autocomplete. It needs its own bug. :-)

Comment 9

[:Gijs (he/him)]

Comment on attachment 8504506 [details] [diff] [review]
asyncresolve can throw,

Approval Request Comment
[Feature/regressing bug #]: maybe bug 494092? I wonder if this is older and needs beta uplift, really - but let's start with aurora
[User impact if declined]: spurious error messages in the console
[Describe test coverage new/current, TBPL]: nope
[Risks and why]: very very very very low
[String/UUID change made/needed]: nope

Comment 10

[Andrei Vaida [:avaida]]

(In reply to :Gijs Kruitbosch from comment #8)
> That's a totally different part of the code to do with the urlbar
> autocomplete. It needs its own bug. :-)

Thanks for the feedback, I'll file a separate bug for it. Updating flags accordingly.

Comment 11

[Andrei Vaida [:avaida]]

(In reply to Andrei Vaida, QA [:avaida] from comment #10)
> I'll file a separate bug for it.
There seems to be an older bug already logged on this matter, Bug 445606. I looked for a regression range but that error is thrown on way earlier builds as well (e.g. 2011-02-01 nightly build), so it's not a recent regression.

Comment 12

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/64eb7cf0d472

Comment 13

[:Gijs (he/him)]

(In reply to :Gijs Kruitbosch from comment #9)
> Comment on attachment 8504506 [details] [diff] [review]
> asyncresolve can throw,
> 
> Approval Request Comment
> [Feature/regressing bug #]: maybe bug 494092? I wonder if this is older and
> needs beta uplift, really - but let's start with aurora

Can't reproduce in 34 beta, so I guess this is really 35+-specific. Andrei, can you confirm that?

Comment 14

[Andrei Vaida [:avaida]]

(In reply to :Gijs Kruitbosch from comment #13)
> Can't reproduce in 34 beta, so I guess this is really 35+-specific. Andrei,
> can you confirm that?

Indeed, I'm not seeing this particular error (Comment 0) on Firefox 34.0b2 (Build ID: 20141020184313). Confirmed on Ubuntu 12.04 LTS 32-bit, Windows 7 64-bit and Mac OS X 10.9.5.

Comment 15

[Andrei Vaida [:avaida]]

Verified fixed on Aurora 35.0a2 (2014-10-21) as well, using Ubuntu 12.04 LTS 32-bit, Windows 7 64-bit and Mac OS X 10.8.5.