Allows extension-devtools api to survive page navigation.

Allows extension-devtools api to survive page navigation.
The problem was that the ExtensionDevtoolsBridge maintained explicit knowledge of the RenderViewHost is was monitoring. This would be invalidated on page navigations where the inspected RVH would be swapped out from underneath it.

This patch removes knowledge of RenderViewHost from the ExtensionDevToolsBridge and makes sure it appropriately unregisters itself.

Patch by: Jaime Yap (jaimeyap@google.com)

TEST=none 
BUG=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=30840

[pfeldman, 2009-11-02 21:59:05]

http://codereview.chromium.org/348042/diff/3001/4003
File chrome/browser/debugger/devtools_manager.h (right):

http://codereview.chromium.org/348042/diff/3001/4003#newcode44
Line 44: void UnregisterDevToolsClientHost(DevToolsClientHost* client_host);
You should be able to call DevToolsClientHost::NotifyCloseListener() on self
instead.

[jaimeyap, 2009-11-02 22:20:15]

http://codereview.chromium.org/348042/diff/3001/4003
File chrome/browser/debugger/devtools_manager.h (right):

http://codereview.chromium.org/348042/diff/3001/4003#newcode44
Line 44: void UnregisterDevToolsClientHost(DevToolsClientHost* client_host);
On 2009/11/02 21:59:06, pfeldman wrote:
> You should be able to call DevToolsClientHost::NotifyCloseListener() on self
> instead.

Done.

[jaimeyap, 2009-11-03 15:32:18]

ping

On 2009/11/02 22:20:15, jaimeyap wrote:
> http://codereview.chromium.org/348042/diff/3001/4003
> File chrome/browser/debugger/devtools_manager.h (right):
> 
> http://codereview.chromium.org/348042/diff/3001/4003#newcode44
> Line 44: void UnregisterDevToolsClientHost(DevToolsClientHost* client_host);
> On 2009/11/02 21:59:06, pfeldman wrote:
> > You should be able to call DevToolsClientHost::NotifyCloseListener() on self
> > instead.
> 
> Done.

[pfeldman, 2009-11-03 17:05:04]

LGTM (now that the tool is alive)

[jaimeyap, 2009-11-03 18:49:10]

@pfeldman: Thanks for the review.

Now I must ask for the obligatory "can someone land this for me" since I don't
have commit access :).

On 2009/11/03 17:05:04, pfeldman wrote:
> LGTM (now that the tool is alive)

[jamesr1, 2009-11-03 19:10:56]

r30840.

On Tue, Nov 3, 2009 at 11:49 AM, <jaimeyap@google.com> wrote:

> @pfeldman: Thanks for the review.
>
> Now I must ask for the obligatory "can someone land this for me" since I
> don't
> have commit access :).
>
>
> On 2009/11/03 17:05:04, pfeldman wrote:
>
>> LGTM (now that the tool is alive)
>>
>
>
>
> http://codereview.chromium.org/348042
>