Fix the commit type for out-of-process iframes.

Fix the commit type for out-of-process iframes.

Passes a has_committed_real_load param to the renderer to let the
(possibly new) FrameLoader know if there have been commits in that
frame in the past, perhaps in another process.

Also fixes broken invariants in NavigationController and tests.

Only affects --site-per-process mode.

BUG=464014
TEST=Navigation tests still pass.

Committed: https://crrev.com/62109d9af7c8db2c1bcaef715f04ca92b4d59746
Cr-Commit-Position: refs/heads/master@{#331712}

[Charlie Reis, 2015-05-27 17:26:22]

creis@chromium.org changed reviewers:
+ avi@chromium.org, japhet@chromium.org

[Charlie Reis, 2015-05-27 17:26:24]

Avi: Can you take a look?  I'm down to one failing test
(SitePerProcessDevToolsBrowserTest.CrossSiteIframeAgentHost), which I'm
currently investigating.  Might make sense to get the review started, though.

Most of the CL is either plumbing the new has_committed_real_load parameter or
fixing broken assumptions once we classify OOPIF commits as auto-subframe.  I've
forked some of them out into smaller CLs to land first.

Nate: Can you give a sanity check in render_frame_impl.cc?

https://codereview.chromium.org/1148953014/diff/140001/content/browser/site_p...
File content/browser/site_per_process_browsertest.cc (right):

https://codereview.chromium.org/1148953014/diff/140001/content/browser/site_p...
content/browser/site_per_process_browsertest.cc:1068: EXPECT_EQ(0U,
site_instance_c->active_frame_count());
This is https://codereview.chromium.org/1155393003/.

https://codereview.chromium.org/1148953014/diff/140001/content/renderer/rende...
File content/renderer/render_frame_impl.cc (right):

https://codereview.chromium.org/1148953014/diff/140001/content/renderer/rende...
content/renderer/render_frame_impl.cc:4413: frame_->setCommittedFirstRealLoad();
japhet: Can you sanity check this part?  I'm basically passing in whether the
frame has already committed a real load in any process, so that we can advance
the (possibly newly created) FrameLoader if needed.

We should get most of the rest for free.

https://codereview.chromium.org/1148953014/diff/140001/content/test/test_fram...
File content/test/test_frame_navigation_observer.h (right):

https://codereview.chromium.org/1148953014/diff/140001/content/test/test_fram...
content/test/test_frame_navigation_observer.h:47: void
DidCommitProvisionalLoadForFrame(
This is https://codereview.chromium.org/1149073004/.

[Charlie Reis, 2015-05-27 19:16:52]

SitePerProcessDevToolsBrowserTest.CrossSiteIframeAgentHost should be fixed in
patch 9.

[Avi (use Gerrit), 2015-05-27 19:26:05]

LGTM otherwise.

https://codereview.chromium.org/1148953014/diff/160001/content/browser/frame_...
File content/browser/frame_host/navigation_controller_impl.cc (right):

https://codereview.chromium.org/1148953014/diff/160001/content/browser/frame_...
content/browser/frame_host/navigation_controller_impl.cc:844: details->type =
new_type;
One way of doing it... :)

You will also need to force ignore_mismatch to true, so you don't trigger the
[D]CHECK during all this mishegas.

[Charlie Reis, 2015-05-27 19:31:19]

Thanks!  I'll wait for the other CLs to land and rebase this before landing it.

Nate, let me know what you think about the render_frame_impl.cc part.

https://codereview.chromium.org/1148953014/diff/160001/content/browser/frame_...
File content/browser/frame_host/navigation_controller_impl.cc (right):

https://codereview.chromium.org/1148953014/diff/160001/content/browser/frame_...
content/browser/frame_host/navigation_controller_impl.cc:844: details->type =
new_type;
On 2015/05/27 19:26:05, Avi wrote:
> One way of doing it... :)

Yeah, I started to fix the old one in patch set 1, but it was kind of a mess. 
Seems like you're close enough to the switch that we can start using the new one
in --site-per-process.  :)

> You will also need to force ignore_mismatch to true, so you don't trigger the
> [D]CHECK during all this mishegas.

The CHECK will never fail in --site-per-process, since this line made the two
values equal.  But you're right that the crash keys aren't needed either, so
I'll add that.

[Avi (use Gerrit), 2015-05-27 19:34:15]

https://codereview.chromium.org/1148953014/diff/160001/content/browser/frame_...
File content/browser/frame_host/navigation_controller_impl.cc (right):

https://codereview.chromium.org/1148953014/diff/160001/content/browser/frame_...
content/browser/frame_host/navigation_controller_impl.cc:844: details->type =
new_type;
On 2015/05/27 19:31:19, Charlie Reis wrote:
> The CHECK will never fail in --site-per-process, since this line made the two
> values equal.

Duh, yeah. My brain is running on the local track today.

[Nate Chapin, 2015-05-27 19:47:40]

render_frame_impl.cc changes LGTM

[Charlie Reis, 2015-05-27 21:38:05]

The CQ bit was checked by creis@chromium.org

[Charlie Reis, 2015-05-27 21:38:05]

The patchset sent to the CQ was uploaded after l-g-t-m from avi@chromium.org,
japhet@chromium.org
Link to the patchset: https://codereview.chromium.org/1148953014/#ps200001
(title: "Rebase for classifier revert")

[commit-bot: I haz the power, 2015-05-27 21:38:46]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1148953014/200001

[Charlie Reis, 2015-05-27 21:41:24]

creis@chromium.org changed reviewers:
+ nasko@chromium.org

[Charlie Reis, 2015-05-27 21:41:25]

Nasko, can you review frame_messages.h for IPC owners?

[nasko, 2015-05-27 21:47:16]

IPC LGTM

[commit-bot: I haz the power, 2015-05-28 00:23:02]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-28 00:23:06]

Try jobs failed on following builders:
  mac_chromium_rel_ng on tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_rel_...)

[nasko, 2015-05-28 00:31:17]

The CQ bit was checked by nasko@chromium.org

[nasko, 2015-05-28 00:31:52]

The CQ bit was unchecked by nasko@chromium.org

[nasko, 2015-05-28 00:32:07]

The CQ bit was checked by nasko@chromium.org

[commit-bot: I haz the power, 2015-05-28 00:32:23]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1148953014/200001

[commit-bot: I haz the power, 2015-05-28 01:28:52]

Committed patchset #11 (id:200001)

[commit-bot: I haz the power, 2015-05-28 01:30:23]

Patchset 11 (id:??) landed as
https://crrev.com/62109d9af7c8db2c1bcaef715f04ca92b4d59746
Cr-Commit-Position: refs/heads/master@{#331712}

[Cait (Slow), 2015-05-28 14:34:04]

A revert of this CL (patchset #11 id:200001) has been created in
https://codereview.chromium.org/1157373002/ by caitkp@chromium.org.

The reason for reverting is: Broke SitesPerProcess Tests on Mac OS 10.9

https://build.chromium.org/p/chromium.mac/builders/Mac10.9%20Tests/builds/2330

Stack trace:
SitePerProcessBrowserTest.RFPHDestruction (run #1):
[ RUN      ] SitePerProcessBrowserTest.RFPHDestruction
[2448:1287:0527/200828:279060882432:WARNING:vt_video_decode_accelerator.cc(198)]
Failed to create hardware VideoToolbox session. Hardware accelerated video
decoding will be disabled.
BrowserTestBase signal handler received SIGTERM. Backtrace:
0   content_browsertests                0x000000010566a983
base::debug::StackTrace::StackTrace() + 19
1   content_browsertests                0x0000000105432fa1 content::(anonymous
namespace)::DumpStackTraceSignalHandler(int) + 65
2   libsystem_platform.dylib            0x00007fff924fd5aa _sigtramp + 26
3   ???                                 0x0000000000000010 0x0 + 16
4   CoreFoundation                      0x00007fff94b28f15
__CFRunLoopServiceMachPort + 181
5   CoreFoundation                      0x00007fff94b28539 __CFRunLoopRun + 1161
6   CoreFoundation                      0x00007fff94b27e75 CFRunLoopRunSpecific
+ 309
7   HIToolbox                           0x00007fff8dc8ca0d
RunCurrentEventLoopInMode + 226
8   HIToolbox                           0x00007fff8dc8c7b7
ReceiveNextEventCommon + 479
9   HIToolbox                           0x00007fff8dc8c5bc
_BlockUntilNextEventMatchingListInModeWithFilter + 65
10  AppKit                              0x00007fff8b3ca24e _DPSNextEvent + 1434
11  AppKit                              0x00007fff8b3c989b -[NSApplication
nextEventMatchingMask:untilDate:inMode:dequeue:] + 122
12  AppKit                              0x00007fff8b3bd99c -[NSApplication run]
+ 553
13  content_browsertests                0x0000000105662681
base::MessagePumpNSApplication::DoRun(base::MessagePump::Delegate*) + 321
14  content_browsertests                0x0000000105661e2c
base::MessagePumpCFRunLoopBase::Run(base::MessagePump::Delegate*) + 92
15  content_browsertests                0x0000000105699eb3 base::RunLoop::Run()
+ 99
16  content_browsertests                0x00000001054416df
content::RunThisRunLoop(base::RunLoop*) + 79
17  content_browsertests                0x0000000105442570
content::NavigateFrameToURL(content::FrameTreeNode*, GURL const&) + 112
18  content_browsertests                0x000000010508b455
content::SitePerProcessBrowserTest_RFPHDestruction_Test::RunTestOnMainThread() +
325
19  content_browsertests                0x00000001053b1d7d
content::ContentBrowserTest::RunTestOnMainThreadLoop() + 205
20  content_browsertests                0x0000000105432d56
content::BrowserTestBase::ProxyRunTestOnMainThreadLoop() + 262
21  content_browsertests                0x00000001053d84a2
content::ShellBrowserMainParts::PreMainMessageLoopRun() + 210
22  content_browsertests                0x0000000107ae36ef
content::BrowserMainLoop::PreMainMessageLoopRun() + 143
23  content_browsertests                0x0000000107d670e7
content::StartupTaskRunner::RunAllTasksNow() + 39
24  content_browsertests                0x0000000107ae1d9e
content::BrowserMainLoop::CreateStartupTasks() + 718
25  content_browsertests                0x0000000107ae5b55
content::BrowserMainRunnerImpl::Initialize(content::MainFunctionParams const&) +
405
26  content_browsertests                0x00000001053d7ee9
ShellBrowserMain(content::MainFunctionParams const&,
scoped_ptr\u003Ccontent::BrowserMainRunner,
base::DefaultDeleter\u003Ccontent::BrowserMainRunner> > const&) + 25
27  content_browsertests                0x00000001053c14ce
content::ShellMainDelegate::RunProcess(std::string const&,
content::MainFunctionParams const&) + 110
28  content_browsertests                0x0000000107a70f35
content::RunNamedProcessTypeMain(std::string const&, content::MainFunctionParams
const&, content::ContentMainDelegate*) + 293
29  content_browsertests                0x0000000107a715bc
content::ContentMainRunnerImpl::Run() + 92
30  content_browsertests                0x0000000107a70c86
content::ContentMain(content::ContentMainParams const&) + 54
31  content_browsertests                0x00000001054329bd
content::BrowserTestBase::SetUp() + 765
32  content_browsertests                0x00000001053b1c22
content::ContentBrowserTest::SetUp() + 322
33  content_browsertests                0x000000010556c66f testing::Test::Run()
+ 319
34  content_browsertests                0x000000010556d2a4
testing::TestInfo::Run() + 436
35  content_browsertests                0x000000010556d733
testing::TestCase::Run() + 467
36  content_browsertests                0x0000000105573c7b
testing::internal::UnitTestImpl::RunAllTests() + 827
37  content_browsertests                0x0000000105573911
testing::UnitTest::Run() + 257
38  content_browsertests                0x0000000105453acb
base::TestSuite::Run() + 171
39  content_browsertests                0x00000001053b28aa
content::ContentTestLauncherDelegate::RunTestSuite(int, char**) + 42
40  content_browsertests                0x000000010543e0d1
content::LaunchTests(content::TestLauncherDelegate*, int, int, char**) + 321
41  content_browsertests                0x00000001053b284a main + 74
42  content_browsertests                0x0000000104ed6234 start + 52
43  ???                                 0x0000000000000009 0x0 + .

[zhaoqin, 2015-05-28 15:15:44]

On 2015/05/28 14:34:04, Cait wrote:
> A revert of this CL (patchset #11 id:200001) has been created in
> https://codereview.chromium.org/1157373002/ by mailto:caitkp@chromium.org.
> 
> The reason for reverting is: Broke SitesPerProcess Tests on Mac OS 10.9
> 
> https://build.chromium.org/p/chromium.mac/builders/Mac10.9%20Tests/builds/2330
> 
> Stack trace:
> SitePerProcessBrowserTest.RFPHDestruction (run #1):
> [ RUN      ] SitePerProcessBrowserTest.RFPHDestruction
>
[2448:1287:0527/200828:279060882432:WARNING:vt_video_decode_accelerator.cc(198)]
> Failed to create hardware VideoToolbox session. Hardware accelerated video
> decoding will be disabled.
> BrowserTestBase signal handler received SIGTERM. Backtrace:
> 0   content_browsertests                0x000000010566a983
> base::debug::StackTrace::StackTrace() + 19
> 1   content_browsertests                0x0000000105432fa1 content::(anonymous
> namespace)::DumpStackTraceSignalHandler(int) + 65
> 2   libsystem_platform.dylib            0x00007fff924fd5aa _sigtramp + 26
> 3   ???                                 0x0000000000000010 0x0 + 16
> 4   CoreFoundation                      0x00007fff94b28f15
> __CFRunLoopServiceMachPort + 181
> 5   CoreFoundation                      0x00007fff94b28539 __CFRunLoopRun +
1161
> 6   CoreFoundation                      0x00007fff94b27e75
CFRunLoopRunSpecific
> + 309
> 7   HIToolbox                           0x00007fff8dc8ca0d
> RunCurrentEventLoopInMode + 226
> 8   HIToolbox                           0x00007fff8dc8c7b7
> ReceiveNextEventCommon + 479
> 9   HIToolbox                           0x00007fff8dc8c5bc
> _BlockUntilNextEventMatchingListInModeWithFilter + 65
> 10  AppKit                              0x00007fff8b3ca24e _DPSNextEvent +
1434
> 11  AppKit                              0x00007fff8b3c989b -[NSApplication
> nextEventMatchingMask:untilDate:inMode:dequeue:] + 122
> 12  AppKit                              0x00007fff8b3bd99c -[NSApplication
run]
> + 553
> 13  content_browsertests                0x0000000105662681
> base::MessagePumpNSApplication::DoRun(base::MessagePump::Delegate*) + 321
> 14  content_browsertests                0x0000000105661e2c
> base::MessagePumpCFRunLoopBase::Run(base::MessagePump::Delegate*) + 92
> 15  content_browsertests                0x0000000105699eb3
base::RunLoop::Run()
> + 99
> 16  content_browsertests                0x00000001054416df
> content::RunThisRunLoop(base::RunLoop*) + 79
> 17  content_browsertests                0x0000000105442570
> content::NavigateFrameToURL(content::FrameTreeNode*, GURL const&) + 112
> 18  content_browsertests                0x000000010508b455
> content::SitePerProcessBrowserTest_RFPHDestruction_Test::RunTestOnMainThread()
+
> 325
> 19  content_browsertests                0x00000001053b1d7d
> content::ContentBrowserTest::RunTestOnMainThreadLoop() + 205
> 20  content_browsertests                0x0000000105432d56
> content::BrowserTestBase::ProxyRunTestOnMainThreadLoop() + 262
> 21  content_browsertests                0x00000001053d84a2
> content::ShellBrowserMainParts::PreMainMessageLoopRun() + 210
> 22  content_browsertests                0x0000000107ae36ef
> content::BrowserMainLoop::PreMainMessageLoopRun() + 143
> 23  content_browsertests                0x0000000107d670e7
> content::StartupTaskRunner::RunAllTasksNow() + 39
> 24  content_browsertests                0x0000000107ae1d9e
> content::BrowserMainLoop::CreateStartupTasks() + 718
> 25  content_browsertests                0x0000000107ae5b55
> content::BrowserMainRunnerImpl::Initialize(content::MainFunctionParams const&)
+
> 405
> 26  content_browsertests                0x00000001053d7ee9
> ShellBrowserMain(content::MainFunctionParams const&,
> scoped_ptr\u003Ccontent::BrowserMainRunner,
> base::DefaultDeleter\u003Ccontent::BrowserMainRunner> > const&) + 25
> 27  content_browsertests                0x00000001053c14ce
> content::ShellMainDelegate::RunProcess(std::string const&,
> content::MainFunctionParams const&) + 110
> 28  content_browsertests                0x0000000107a70f35
> content::RunNamedProcessTypeMain(std::string const&,
content::MainFunctionParams
> const&, content::ContentMainDelegate*) + 293
> 29  content_browsertests                0x0000000107a715bc
> content::ContentMainRunnerImpl::Run() + 92
> 30  content_browsertests                0x0000000107a70c86
> content::ContentMain(content::ContentMainParams const&) + 54
> 31  content_browsertests                0x00000001054329bd
> content::BrowserTestBase::SetUp() + 765
> 32  content_browsertests                0x00000001053b1c22
> content::ContentBrowserTest::SetUp() + 322
> 33  content_browsertests                0x000000010556c66f
testing::Test::Run()
> + 319
> 34  content_browsertests                0x000000010556d2a4
> testing::TestInfo::Run() + 436
> 35  content_browsertests                0x000000010556d733
> testing::TestCase::Run() + 467
> 36  content_browsertests                0x0000000105573c7b
> testing::internal::UnitTestImpl::RunAllTests() + 827
> 37  content_browsertests                0x0000000105573911
> testing::UnitTest::Run() + 257
> 38  content_browsertests                0x0000000105453acb
> base::TestSuite::Run() + 171
> 39  content_browsertests                0x00000001053b28aa
> content::ContentTestLauncherDelegate::RunTestSuite(int, char**) + 42
> 40  content_browsertests                0x000000010543e0d1
> content::LaunchTests(content::TestLauncherDelegate*, int, int, char**) + 321
> 41  content_browsertests                0x00000001053b284a main + 74
> 42  content_browsertests                0x0000000104ed6234 start + 52
> 43  ???                                 0x0000000000000009 0x0 + .

It also break many memory fyi bots
http://build.chromium.org/p/chromium.memory.fyi/builders/Linux%20MSan%20Tests...
http://build.chromium.org/p/chromium.memory.fyi/builders/Linux%20ChromeOS%20M...
http://build.chromium.org/p/chromium.memory.fyi/builders/Windows%20Content%20...