Avoid fetching RFH from navigation handle for not committed navigations.

Avoid fetching RFH from nav handle for not committed navigations.

For devtool use case there is only a need to check if hosts match,
so compareing frame tree node id instead of raw pointers will give
same result.

BUG=697991
Review-Url: https://codereview.chromium.org/2730873002
Cr-Commit-Position: refs/heads/master@{#454849}
Committed: https://chromium.googlesource.com/chromium/src/+/9b5ea61b97315d1d028f440d6dc7defed009cf66

[mharanczyk, 2017-03-03 10:22:42]

Description was changed from

==========
Avoid fetching RFH from navigation handle for not committed navigations.

FOr devtool use case there is only a need to check if hosts match,
so compareing frame tree node id instead of raw pointers will give
same result.

BUG=697991
==========

to

==========
Avoid fetching RFH from nav handle for not committed navigations.

For devtool use case there is only a need to check if hosts match,
so compareing frame tree node id instead of raw pointers will give
same result.

BUG=697991
==========

[mharanczyk, 2017-03-03 10:47:03]

mharanczyk@opera.com changed reviewers:
+ dgozman@chromium.org, jam@chromium.org

[mharanczyk, 2017-03-03 10:47:04]

Please take a look.

To be honest I don't really know how much of a workaround is this solution. It
uses almost exactly same data for comparison
(NavigationHandleImpl::frame_tree_node_ or its field) only difference being that
one function is allowed to be called earlier than other due to CHECK. I have
found other places in code that calls GetFrameTreeNodeId in earlier stages of
navigation (for example DidStartNavigation) so I guess it is fine to call it
here as well, at least for now.

[mharanczyk, 2017-03-03 12:12:42]

To be clear: I made this change under the assumption that fetching RFH from not
committed navigation is wrong because you could call EvilFunction on it that
will break stuff, but GetFrameTreeNodeId is one of the safe functions to call in
this situation.

[jam, 2017-03-03 17:17:05]

lgtm

[dgozman, 2017-03-03 18:21:29]

lgtm. Thanks!

[mharanczyk, 2017-03-06 10:29:57]

The CQ bit was checked by mharanczyk@opera.com

[commit-bot: I haz the power, 2017-03-06 10:30:12]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-06 11:15:25]

CQ is committing da patch.

Bot data: {"patchset_id": 1, "attempt_start_ts": 1488796197913180, "parent_rev":
"122c4da76a70443fe965d1413741a79416a6e763", "commit_rev":
"9b5ea61b97315d1d028f440d6dc7defed009cf66"}

[commit-bot: I haz the power, 2017-03-06 11:16:00]

Description was changed from

==========
Avoid fetching RFH from nav handle for not committed navigations.

For devtool use case there is only a need to check if hosts match,
so compareing frame tree node id instead of raw pointers will give
same result.

BUG=697991
==========

to

==========
Avoid fetching RFH from nav handle for not committed navigations.

For devtool use case there is only a need to check if hosts match,
so compareing frame tree node id instead of raw pointers will give
same result.

BUG=697991

Review-Url: https://codereview.chromium.org/2730873002
Cr-Commit-Position: refs/heads/master@{#454849}
Committed:
https://chromium.googlesource.com/chromium/src/+/9b5ea61b97315d1d028f440d6dc7...
==========

[commit-bot: I haz the power, 2017-03-06 11:16:01]

Committed patchset #1 (id:1) as
https://chromium.googlesource.com/chromium/src/+/9b5ea61b97315d1d028f440d6dc7...