Revert of PlzNav: Fix NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…

Revert of PlzNav: Fix NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga… (patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
>
> This test is failing in PlzNav because
> https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_controller_impl_browsertest.cc?q=%22Simulate+the+user+hitting+Enter%22&sq=package:chromium&l=5943&dr=C
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
>
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
>
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
>
> *That* is because here
> https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/FrameLoader.cpp?rcl=0&l=839
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
>
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
>
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
>
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103, 475027, 536102
>
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103, 475027, 536102, 664319
Committed: https://crrev.com/fe5ea3f7cb872ea6fa2e560cfefc9185393a8cb2
Cr-Commit-Position: refs/heads/master@{#431896}

[scottmg, 2016-11-14 17:59:47]

The CQ bit was checked by scottmg@chromium.org

[scottmg, 2016-11-14 17:59:48]

Created Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…

[scottmg, 2016-11-14 18:00:16]

Description was changed from

==========
Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
(patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
> 
> This test is failing in PlzNav because
>
https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_co...
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
> 
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
> 
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
> 
> *That* is because here
>
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/Fr...
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
> 
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
> 
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
> 
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103,475027, 536102
> 
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103,475027, 536102
==========

to

==========
Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
(patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
> 
> This test is failing in PlzNav because
>
https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_co...
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
> 
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
> 
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
> 
> *That* is because here
>
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/Fr...
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
> 
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
> 
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
> 
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103,475027, 536102
> 
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103,475027, 536102,664319
==========

[commit-bot: I haz the power, 2016-11-14 18:00:34]

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, 2016-11-14 19:20:39]

Description was changed from

==========
Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
(patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
> 
> This test is failing in PlzNav because
>
https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_co...
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
> 
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
> 
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
> 
> *That* is because here
>
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/Fr...
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
> 
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
> 
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
> 
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103,475027, 536102
> 
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103,475027, 536102,664319
==========

to

==========
Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
(patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
> 
> This test is failing in PlzNav because
>
https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_co...
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
> 
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
> 
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
> 
> *That* is because here
>
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/Fr...
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
> 
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
> 
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
> 
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103,475027, 536102
> 
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103,475027, 536102,664319
==========

[commit-bot: I haz the power, 2016-11-14 19:20:45]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-11-14 19:35:16]

Description was changed from

==========
Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
(patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
> 
> This test is failing in PlzNav because
>
https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_co...
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
> 
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
> 
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
> 
> *That* is because here
>
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/Fr...
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
> 
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
> 
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
> 
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103,475027, 536102
> 
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103,475027, 536102,664319
==========

to

==========
Revert of PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
(patchset #21 id:500001 of https://codereview.chromium.org/2381503003/ )

Reason for revert:
https://crbug.com/664319.

Original issue's description:
> PlzNav: Fix
NavigationControllerBrowserTest.EnsureSamePageNavigationUpdatesFrameNaviga…
>
> This test is failing in PlzNav because
>
https://cs.chromium.org/chromium/src/content/browser/frame_host/navigation_co...
> is creating a new navigation entry, so the GoBack() goes to the wrong
> place.
>
> It's doing that because
> FrameHostMsg_DidCommitProvisionalLoad_Params.did_create_new_entry is
> true.
>
> That in turn is because SendDidCommitProvisionalLoad is getting a commit
> type of WebStandardCommit (not WebHistoryInertCommit like it does in
> non-PlzNav).
>
> *That* is because here
>
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/loader/Fr...
> the url being checked is the already-redirected one in the PlzNavigate
> case (sub.a.com), but m_documentLoader->urlForHistory() is the
> pre-redirected one (a.com/server-redirect?sub...).
>
> So, this works in non-PlzNav because Blink tells us that it should
> replace the current entry (i.e. it's like a reload).
>
> Instead of relying on Blink to determine that, check if the render frame
> node's last committed url is the original url we navigated to there, and
> that we're not just mucking with history navigation, in which case we
> convert the normal load to a reload.
>
> R=nasko@chromium.org,clamy@chromium.org
> BUG=630103,475027, 536102
>
> Committed: https://crrev.com/28bbbb16194c7337e22c3e0bda752ae6b4b681cf
> Cr-Commit-Position: refs/heads/master@{#429302}

TBR=nasko@chromium.org,clamy@chromium.org,creis@chromium.org,boliu@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=630103,475027, 536102,664319

Committed: https://crrev.com/fe5ea3f7cb872ea6fa2e560cfefc9185393a8cb2
Cr-Commit-Position: refs/heads/master@{#431896}
==========

[commit-bot: I haz the power, 2016-11-14 19:35:18]

Patchset 1 (id:??) landed as
https://crrev.com/fe5ea3f7cb872ea6fa2e560cfefc9185393a8cb2
Cr-Commit-Position: refs/heads/master@{#431896}