cc: Add an integration test for what scale we draw with after pinching

cc: Add an integration test for what scale we draw with after pinching

This is a LayerTreeTest that does:
1) Pinch in and end the pinch
2) Wait to draw fully at the new scale (2.2x)
3) Pinch out but hold onto the pinch
4) Wait to draw fully at the new scale (1.1x)
5) Prevent new tiles from being ready
6) End the pinch
7) Let one draw happen, which will let the raster scale change (still
   at 1.1x)
8) Let tiles become ready
9) Ensure we draw tiles with the ideal scale

Step 9 wasn't happening for the bug this is testing.

R=vmpstr
BUG=427423

[danakj, 2014-10-28 21:14:07]

https://codereview.chromium.org/690493002/diff/1/cc/test/fake_content_layer_c...
File cc/test/fake_content_layer_client.cc (right):

https://codereview.chromium.org/690493002/diff/1/cc/test/fake_content_layer_c...
cc/test/fake_content_layer_client.cc:45: bool red = true;
This makes every tile !solidcolor

[danakj, 2014-10-29 14:36:51]

This test seems to expose this cc weakptr bug more. jbroman@ claims to have seen
similar on layout tests which may be this same instance of a bug:

[18579:18579:1029/103538:1524935088547:FATAL:weak_ptr.cc(20)] Check failed:
sequence_checker_.CalledOnValidSequencedThread() || HasOneRef(). WeakPtrs must
be invalidated on the same sequenced thread.
#0 0x7f3444dc44ce base::debug::StackTrace::StackTrace()
#1 0x7f3444e5ebf5 logging::LogMessage::~LogMessage()
#2 0x7f3444e823ba base::internal::WeakReference::Flag::Invalidate()
#3 0x7f3444e826b6 base::internal::WeakReferenceOwner::Invalidate()
#4 0x7f3444e82669 base::internal::WeakReferenceOwner::~WeakReferenceOwner()
#5 0x7f34451ff9ed base::WeakPtrFactory<>::~WeakPtrFactory()
#6 0x7f3445209a45 cc::DelayedUniqueNotifier::~DelayedUniqueNotifier()
#7 0x7f344555734b cc::ThreadProxy::CompositorThreadOnly::~CompositorThreadOnly()
#8 0x7f34455576d3 cc::ThreadProxy::~ThreadProxy()
#9 0x000000da9a85 cc::ThreadProxyForTest::~ThreadProxyForTest()
#10 0x000000da9aa9 cc::ThreadProxyForTest::~ThreadProxyForTest()
#11 0x7f34454e9ed2 base::DefaultDeleter<>::operator()()
#12 0x7f34454ecf74 base::internal::scoped_ptr_impl<>::~scoped_ptr_impl()
#13 0x7f34454dd815 scoped_ptr<>::~scoped_ptr()
#14 0x7f34454d5270 cc::LayerTreeHost::~LayerTreeHost()
#15 0x000000da9f53 cc::LayerTreeHostForTesting::~LayerTreeHostForTesting()
#16 0x000000da9f79 cc::LayerTreeHostForTesting::~LayerTreeHostForTesting()
#17 0x0000004ba6e2 base::DefaultDeleter<>::operator()()

I'm going to try figure out what we're doing wrong with weakptrs.

[danakj, 2014-10-29 19:28:21]

I'm going to close this and add the test with the "real fix" to the bug.