Reland r152781 - "Fix a history bug where a reload is given a standard load's history entry."

Reland r152781 - "Fix a history bug where a reload is given a standard load's history entry."

This reverts commit d1dc33d638f4657d4b4563de30d8f2bc089f8850.

BUG=241581
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=152974

[Nate Chapin, 2013-06-24 19:23:48]

This patch was reverted because of a failing android test. Unfortunately, I
can't get the relevant android bot to accept my patch, so this patch is
speculative (though I'm about 90% confident it's right, based on the failure).

The failure:
http://build.chromium.org/p/chromium.webkit/builders/Android%20Tests%20%28dbg...

The test (ModalDialogTest#testBeforeUnloadDialog) checks whether we show a
before unload dialog under a couple different scenarios.

I believe it failed because I changed the point at which a new load sets
FrameLoader::m_loadType from just before the before unload dialog check to just
after. This means the before unload dialog check was using the previous load
type instead of the new one.

Is it ok to land it and see what happens?

https://codereview.chromium.org/17519002/diff/1/Source/core/loader/FrameLoade...
File Source/core/loader/FrameLoader.cpp (right):

https://codereview.chromium.org/17519002/diff/1/Source/core/loader/FrameLoade...
Source/core/loader/FrameLoader.cpp:2082: if ((isTargetItem ||
isLoadingMainFrame()) && isBackForwardLoadType(type)) {
This is the only diff from the patch that was reverted.

[abarth-chromium, 2013-06-24 20:04:25]

On 2013/06/24 19:23:48, Nate Chapin wrote:
> Is it ok to land it and see what happens?

Sure.  It's unfortunate that we can't use the try server...  I might let the
gardener know when you're landing this CL.

LGTM

[commit-bot: I haz the power, 2013-06-24 20:06:35]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/japhet@chromium.org/17519002/7001

[commit-bot: I haz the power, 2013-06-24 23:29:33]

Change committed as 152974