Consider "mixed content XHR" as mixed script instead of mixed display.

Consider "mixed content XHR" as mixed script instead of mixed display.

Second attempt.  The first attempt at https://codereview.chromium.org/14619021/
was reverted because it was over-aggressive: it could not distinguish between
raw XHR requests and other RAW (audio/video) request.

This CL adds a new flag to ResourceLoaderOptions to distinguish the two
cases, and check this in ResourceFetcher.cpp.  DocumentThreadableLoader's
sync case is then made to use the new checking code in ResourceFetcher in
place of its own check.

BUG=239540
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=157139

[Tom Sepez, 2013-08-28 21:27:24]

Adam, please review.

[Chris Evans, 2013-08-28 21:29:02]

On 2013/08/28 21:27:24, Tom Sepez wrote:
> Adam, please review.

[FWIW, the first attempt also failed to distinguish (and allow) PPAPI Flash
subresource mixed loads. Looks like you have that covered here]

[Tom Sepez, 2013-08-29 17:59:54]

@japhet - do you have an opinion on whether this would be cleaner if type()
Resource::Raw was split into Resource::ActiveRaw and Resource::PassiveRaw vs.
adding the flag to options?  Thanks.

[Tom Sepez, 2013-08-29 20:38:49]

> @japhet - do you have an opinion on whether this would be cleaner if type()
> Resource::Raw was split into Resource::ActiveRaw and Resource::PassiveRaw vs.
> adding the flag to options?  Thanks.
(My instinct is "no" having tried to see where this flows to ...)

[abarth-chromium, 2013-08-30 06:37:07]

https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
File Source/core/fetch/ResourceFetcher.cpp (right):

https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
Source/core/fetch/ResourceFetcher.cpp:349: }
Should this be "else if" ?  I don't think we ever want to call both of these
functions.

https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
File Source/core/fetch/ResourceLoaderOptions.h (right):

https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
Source/core/fetch/ResourceLoaderOptions.h:95: TreatAsActiveContent
I'd recommend making this a three-state enum.  The extra value would be
TreatAsDefaultForType, which would give you the default value.  Then these two
values would force the check to be either passive or active, regardless of the
type.  We can then add ASSERTs in checkInsecureContent, for example, to make
sure we never get TreatAsDefaultForType for Raw resources.

[Tom Sepez, 2013-08-30 17:09:50]

On 2013/08/30 06:37:07, abarth wrote:
>
https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
> File Source/core/fetch/ResourceFetcher.cpp (right):
> 
>
https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
> Source/core/fetch/ResourceFetcher.cpp:349: }
> Should this be "else if" ?  I don't think we ever want to call both of these
> functions.
> 
Will do.

>
https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
> File Source/core/fetch/ResourceLoaderOptions.h (right):
> 
>
https://codereview.chromium.org/23437013/diff/8001/Source/core/fetch/Resource...
> Source/core/fetch/ResourceLoaderOptions.h:95: TreatAsActiveContent
> I'd recommend making this a three-state enum.  The extra value would be
> TreatAsDefaultForType, which would give you the default value.  Then these two
> values would force the check to be either passive or active, regardless of the
> type.  We can then add ASSERTs in checkInsecureContent, for example, to make
> sure we never get TreatAsDefaultForType for Raw resources.
I like it.

[Tom Sepez, 2013-08-30 17:58:40]

Adam, I stopped short of asserting that all uses of type Raw had to specify an
explicit TreatAs... since only the one path is active, and the others can
default to the old, passive behaviour.  I also added a
TreatAsAlwaysAllowedContent just for the sake of completeness; since we're
adding an option to override the switch(type), we might as well cover all the
cases that could be otherwise selected, even if its not used right now.

[abarth-chromium, 2013-09-03 15:48:45]

lgtm

[commit-bot: I haz the power, 2013-09-03 16:38:58]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tsepez@chromium.org/23437013/26001

[commit-bot: I haz the power, 2013-09-03 17:54:07]

Change committed as 157139