Add CORS-awareness to HTML5 media elements.

Add CORS-awareness to HTML5 media elements.

This CL is dead code until WebMediaPlayerClientImpl starts calling the new 2-arg
load() method, but this needs to land (and roll into WebKit DEPS) before WMPCI
can be updated to call this 2-arg load().

BUG=123369
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=140779

[scherkus (not reviewing), 2012-06-06 00:44:36]

wooooooot

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/androi...
File webkit/media/android/webmediaplayer_android.h (right):

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/androi...
webkit/media/android/webmediaplayer_android.h:59: const WebKit::WebString&
cross_origin_attr);
CrossOriginAttribute cross_origin_attribute instead?

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/buffer...
File webkit/media/buffered_resource_loader.h (right):

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/buffer...
webkit/media/buffered_resource_loader.h:69: enum CrossOriginAttribute {
kUnspecified, kAnonymous, kUseCredentials };
are you following abarth's suggestion re: CORSMode?

the word "attribute" is more of a WebCore/html-ism and shouldn't be present down
here in the depths of media engine code!

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/webmed...
File webkit/media/webmediaplayer_impl.cc (right):

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/webmed...
webkit/media/webmediaplayer_impl.cc:232: DCHECK_EQ(cross_origin_attribute,
Unspecified);
so adding a crossorigin attribute to a media stream <video> in debug build will
crash? that doesn't seem very nice :(

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/webmed...
File webkit/media/webmediaplayer_impl.h (right):

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/webmed...
webkit/media/webmediaplayer_impl.h:116: CrossOriginAttribute
cross_origin_attribute);
is this a WK::COA or a BRL::COA?

I'm assuming a WK one but shouldn't it be prefixed?

[Ami GONE FROM CHROMIUM, 2012-06-06 00:59:28]

Not sure how I managed to screw up PS#2's upload; it should have contained the
CORSMode change already.  Addressed your comments where they weren't already
addressed.

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/webmed...
File webkit/media/webmediaplayer_impl.cc (right):

https://chromiumcodereview.appspot.com/10543007/diff/2001/webkit/media/webmed...
webkit/media/webmediaplayer_impl.cc:232: DCHECK_EQ(cross_origin_attribute,
Unspecified);
On 2012/06/06 00:44:36, scherkus wrote:
> so adding a crossorigin attribute to a media stream <video> in debug build
will
> crash? that doesn't seem very nice :(

Done.

https://chromiumcodereview.appspot.com/10543007/diff/5002/webkit/media/webmed...
File webkit/media/webmediaplayer_impl.h (right):

https://chromiumcodereview.appspot.com/10543007/diff/5002/webkit/media/webmed...
webkit/media/webmediaplayer_impl.h:115: virtual void load(const WebKit::WebURL&
url, CORSMode cors_mode);
It doesn't need to be prefixed b/c it lives inside WMPI, which is-a WK::WMP,
which is what contains the CORSMode type.

[scherkus (not reviewing), 2012-06-06 01:19:58]

LGTM but I think we should ignore cors_mode on non-regular-WMPIs

https://chromiumcodereview.appspot.com/10543007/diff/11001/webkit/media/webme...
File webkit/media/webmediaplayer_impl.cc (right):

https://chromiumcodereview.appspot.com/10543007/diff/11001/webkit/media/webme...
webkit/media/webmediaplayer_impl.cc:228: if (cors_mode == Unspecified &&
isn't the proper behaviour to ignore the attribute because it doesn't make
sense? otherwise we'll end up skipping the entire media stream code path and
attempt to load the media stream URL assuming it's a media file

[Ami GONE FROM CHROMIUM, 2012-06-06 01:28:40]

https://chromiumcodereview.appspot.com/10543007/diff/11001/webkit/media/webme...
File webkit/media/webmediaplayer_impl.cc (right):

https://chromiumcodereview.appspot.com/10543007/diff/11001/webkit/media/webme...
webkit/media/webmediaplayer_impl.cc:228: if (cors_mode == Unspecified &&
On 2012/06/06 01:19:58, scherkus wrote:
> isn't the proper behaviour to ignore the attribute because it doesn't make
> sense? otherwise we'll end up skipping the entire media stream code path and
> attempt to load the media stream URL assuming it's a media file

Done.

[commit-bot: I haz the power, 2012-06-06 01:29:02]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fischman@chromium.org/10543007/14001

[commit-bot: I haz the power, 2012-06-06 01:39:43]

Try job failure for 10543007-14001 (retry) on mac_rel for step "compile"
(clobber build).
It's a second try, previously, step "compile" failed.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[Ami GONE FROM CHROMIUM, 2012-06-06 02:25:31]

Grrr...  Need to land the new enum in WebKit first: http://webk.it/88388

[commit-bot: I haz the power, 2012-06-06 14:49:39]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fischman@chromium.org/10543007/4006

[commit-bot: I haz the power, 2012-06-06 15:12:02]

Try job failure for 10543007-4006 (retry) on mac_rel for step "compile" (clobber
build).
It's a second try, previously, step "compile" failed.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[commit-bot: I haz the power, 2012-06-06 15:17:41]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fischman@chromium.org/10543007/13003

[commit-bot: I haz the power, 2012-06-06 17:01:32]

Change committed as 140779

[rniwa-cr, 2012-06-07 03:34:53]

This patch broke http/tests/media/video-cookie.html:

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=htt...

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog.html?url=%2F...

[rniwa-cr, 2012-06-07 03:39:14]

Filed https://code.google.com/p/chromium/issues/detail?id=131533.