1381771 - fix mochitest-chrome failure in devtools/ when data: URI is unique origin.

Status: RESOLVED FIXED

(Core :: DOM: Security, defect, P2)

Comment 1

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Attachment: Patch.

Comment 2

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Comment on attachment 8887811 [details] [diff] [review]
Patch.

Review of attachment 8887811 [details] [diff] [review]:
-----------------------------------------------------------------

::: devtools/shared/webconsole/test/test_commands_other.html
@@ +78,2 @@
>    info ("Waiting for content window to load");
> +  startTest();

Hi smaug
now I open a http page, the gWin.onload won't be called,
I guess this is because the gWin is not in the same origin(now test_commands_other.html is a chrome test)

However, even I try to do window.opener.postMessage in empty.html, it says "Permission denied to access property postMessage'

Do you have any suggestion to know that empty.html is loaded?

Thanks

Comment 3

[Olli Pettay [:smaug][[e-mail]]]

What is opener here? That kind of security exception sounds odd. Is opener a chrome window and window itself non-chrome window? I mean, what kind of principals they have?

Comment 4

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Attachment: Part 1: fix failures in devtools/

Comment 5

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Attachment: Part 2: use legacy data: URI behavior to run test.

Comment 6

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Hi pbro
I found you turn off r? flag for now, if you are too busy to review this, can you help to forward the r? to right one?

Thanks

Comment 7

[Patrick Brosset <:pbro>]

Comment on attachment 8888666 [details] [diff] [review]
Part 1: fix failures  in devtools/

Review of attachment 8888666 [details] [diff] [review]:
-----------------------------------------------------------------

I'm actually going away later today, I have already turned off review requests to avoid getting a big one I know I can't do in one day. But this is a small one, so I've done it now.
Thanks for working on this.

Comment 8

[Patrick Brosset <:pbro>]

Comment on attachment 8888667 [details] [diff] [review]
Part 2: use legacy data: URI behavior to run test.

Review of attachment 8888667 [details] [diff] [review]:
-----------------------------------------------------------------

Assigning this console-related one to bgrins.

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/83e350c46e94
Part 1: fix mc1 failures in devtools/ when data: URI is unique origin. r=pbro
https://hg.mozilla.org/integration/mozilla-inbound/rev/1f551cfb7a2d
Part 2: use legacy data: URI behavior to run test. r=bgrins

Comment 10

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/83e350c46e94
https://hg.mozilla.org/mozilla-central/rev/1f551cfb7a2d