<webview>: Fix resize.

<webview>: Fix resize.

BrowserPluginHostMsg_UpdateRect_ACK ships BrowserPluginHostMsg_ResizeGuest_Params
in the message. Most of the time this struct is uninitialized, and unused. However,
BrowserPluginGuest::OnResizeGuest is called regardless. This resets resize flags,
and occasionally confuses RenderWidgetHostImpl by calling WasResized indirectly.

On certain websites, this breaks resize, or may cause resize to be delayed until
a mouse movement is performed over the content.

This patch addresses the issue by adding a size_changed field to the struct. All other fields are ignored unless that field is true.

BUG=263027
Test=manually, visit http://www.html5rocks.com/en/tutorials/pointerlock/intro in the Chrome Apps Browser demo

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=213021

[alexst (slow to review), 2013-07-22 19:09:02]

lgtm

[Fady Samuel, 2013-07-22 19:43:33]

+cdn@ for small browser_plugin_messages.h change.

[Cris Neckar, 2013-07-22 19:56:05]

https://codereview.chromium.org/19585004/diff/1/content/browser/browser_plugi...
File content/browser/browser_plugin/browser_plugin_guest.cc (right):

https://codereview.chromium.org/19585004/diff/1/content/browser/browser_plugi...
content/browser/browser_plugin/browser_plugin_guest.cc:1252: if
(!params.size_changed)
I don't understand why you need to send this? If the message does nothing when
it is set couldn't you just as easily not send to message at all in this case?

[Fady Samuel, 2013-07-22 20:20:46]

https://codereview.chromium.org/19585004/diff/1/content/browser/browser_plugi...
File content/browser/browser_plugin/browser_plugin_guest.cc (right):

https://codereview.chromium.org/19585004/diff/1/content/browser/browser_plugi...
content/browser/browser_plugin/browser_plugin_guest.cc:1252: if
(!params.size_changed)
On 2013/07/22 19:56:05, Cris Neckar wrote:
> I don't understand why you need to send this? If the message does nothing when
> it is set couldn't you just as easily not send to message at all in this case?

This method is called from multiple places (not just
BrowserPluginHostMsg_ResizeGuest). The params are piggybacked off multiple
message types. Sometimes the fields of BrowserPluginHostMsg_ResizeGuest_Params
are not populated.

[Cris Neckar, 2013-07-22 20:44:21]

On 2013/07/22 20:20:46, Fady Samuel wrote:
>
https://codereview.chromium.org/19585004/diff/1/content/browser/browser_plugi...
> File content/browser/browser_plugin/browser_plugin_guest.cc (right):
> 
>
https://codereview.chromium.org/19585004/diff/1/content/browser/browser_plugi...
> content/browser/browser_plugin/browser_plugin_guest.cc:1252: if
> (!params.size_changed)
> On 2013/07/22 19:56:05, Cris Neckar wrote:
> > I don't understand why you need to send this? If the message does nothing
when
> > it is set couldn't you just as easily not send to message at all in this
case?
> 
> This method is called from multiple places (not just
> BrowserPluginHostMsg_ResizeGuest). The params are piggybacked off multiple
> message types. Sometimes the fields of BrowserPluginHostMsg_ResizeGuest_Params
> are not populated.

gotcha IPC changes lgtm

[commit-bot: I haz the power, 2013-07-22 20:45:20]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/fsamuel@chromium.org/19585004/1

[commit-bot: I haz the power, 2013-07-23 01:13:33]

Change committed as 213021