Merge lp:~andreas-pokorny/mir/fix-surface-information-not-propagated-to-android-input-stack into lp:mir

[Andreas Pokorny]

Updates the InputWindowInfo found at InputWindowHandles everytime the InputDispatcher tries to querry the object for hit testing.

 So far that information was only updated during creation (because we triggered that in the mir implemantation) and when focus changes occur - that might be the reason why unity8 never experienced these issues. This fix kind of makes updateInfo and the interface of InputWindowInfo even more questionable. Why have those members public? Why update these only under certain circumstances..

This changes updates the InputWindowInfo that is provided to the android InputDispatcher for each surface and application (?) just before those objects are tested against touch positions.

[PS Jenkins bot (ps-jenkins)]

PASSED: Continuous integration, rev:1436
http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-ci/960/
Executed test runs:
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-android-trusty-i386-build/1079
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-clang-trusty-amd64-build/1077
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-trusty-touch/669
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-amd64-ci/692
        deb: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-amd64-ci/692/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-armhf-ci/697
        deb: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-armhf-ci/697/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-builder-trusty-armhf/670
        deb: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-builder-trusty-armhf/670/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-runner-mako/638
    SUCCESS: http://s-jenkins.ubuntu-ci:8080/job/touch-flash-device/4379

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/mir-team-mir-development-branch-ci/960/rebuild

review: Approve (continuous-integration)

[Kevin DuBois (kdub)]

looks good to me

review: Approve

[Alan Griffiths (alan-griffiths)]

LGTM

review: Approve

[Robert Carr (robertcarr)]

+1. Your suspicion is right that WindowInfo and GetInfo is ugly. We inherited it from android input. I think in particular they used it to solve threading issues with the C++ input stack and a WindowInfo implementation on the Java side.

review: Approve

[Daniel van Vugt (vanvugt)]

Verified bug 1268819 is fixed! However bug 1261647 is not (try the test case in the bug description). Unlinking the second bug for now...

review: Approve

[PS Jenkins bot (ps-jenkins)]

FAILED: Autolanding.
More details in the following jenkins job:
http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-autolanding/433/
Executed test runs:
    FAILURE: http://jenkins.qa.ubuntu.com/job/mir-android-trusty-i386-build/1101/console
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-clang-trusty-amd64-build/1099
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-trusty-touch/688
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-amd64-autolanding/338
        deb: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-amd64-autolanding/338/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-armhf-autolanding/339
        deb: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-armhf-autolanding/339/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-builder-trusty-armhf/689
        deb: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-builder-trusty-armhf/689/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-mediumtests-runner-mako/656
    SUCCESS: http://s-jenkins.ubuntu-ci:8080/job/touch-flash-device/4440

review: Needs Fixing (continuous-integration)

[Daniel van Vugt (vanvugt)]

Blocked on bug 1287600 :(

[Andreas Pokorny (andreas-pokorny)]

> Verified bug 1268819 is fixed! However bug 1261647 is not (try the test case
> in the bug description). Unlinking the second bug for now...

ah ok, fingerpaint being fullscreen on start is a bad test subject. Will dig again.

[Alan Griffiths (alan-griffiths)]

bug 1287600 is fixed. top approving

review: Approve

[PS Jenkins bot (ps-jenkins)]

review: Approve (continuous-integration)

=== modified file '3rd_party/android-input/android/frameworks/base/services/input/InputDispatcher.cpp'
--- 3rd_party/android-input/android/frameworks/base/services/input/InputDispatcher.cpp	2014-01-13 06:12:33 +0000
+++ 3rd_party/android-input/android/frameworks/base/services/input/InputDispatcher.cpp	2014-02-28 10:23:51 +0000
@@ -432,6 +432,7 @@
 sp<InputWindowHandle> InputDispatcher::findTouchedWindowAtLocked(int32_t x, int32_t y) {
     sp<InputWindowHandle> foundHandle = NULL;
     mEnumerator->for_each([&](sp<InputWindowHandle> windowHandle) {
+        windowHandle->updateInfo();
         const InputWindowInfo* windowInfo = windowHandle->getInfo();
         int32_t flags = windowInfo->layoutParamsFlags;
 
@@ -1151,6 +1152,7 @@
 
         // Traverse windows from front to back to find touched window and outside targets.
         mEnumerator->for_each([&](sp<InputWindowHandle> const& windowHandle){
+            windowHandle->updateInfo();
             const InputWindowInfo* windowInfo = windowHandle->getInfo();
             int32_t flags = windowInfo->layoutParamsFlags;
 

=== modified file 'tests/acceptance-tests/test_client_input.cpp'
--- tests/acceptance-tests/test_client_input.cpp	2014-02-18 10:33:14 +0000
+++ tests/acceptance-tests/test_client_input.cpp	2014-02-28 10:23:51 +0000
@@ -608,3 +608,41 @@
     launch_client_process(*client_config_1);
     launch_client_process(*client_config_2);
 }
+
+TEST_F(TestClientInput, clients_receive_motion_within_co_ordinate_system_of_window)
+{
+    using namespace ::testing;
+
+    static int const screen_width = 1000;
+    static int const screen_height = 800;
+    static int const client_height = screen_height/2;
+    static int const client_width = screen_width/2;
+    static std::string const test_client = "tc";
+    mtf::CrossProcessSync fence;
+
+    static GeometryMap positions;
+    positions[test_client] = geom::Rectangle{geom::Point{screen_width/2, screen_height/2},
+                                             geom::Size{client_width, client_height}};
+
+    auto server_config = make_event_producing_server(fence, 1,
+         [&](mtf::InputTestingServerConfiguration& server)
+         {
+            server.the_session_container()->for_each([&](std::shared_ptr<msh::Session> const& session) -> void
+            {
+                session->default_surface()->move_to(geom::Point{screen_width/2-40, screen_height/2-80});
+            });
+            server.fake_event_hub->synthesize_event(mis::a_motion_event().with_movement(screen_width/2+40, screen_height/2+90));
+        }, positions, DepthMap());
+    launch_server_process(*server_config);
+
+    auto client = make_event_expecting_client(test_client, fence,
+         [&](MockHandler& handler, mt::WaitCondition& events_received)
+         {
+             InSequence seq;
+             EXPECT_CALL(handler, handle_input(mt::HoverEnterEvent())).Times(1);
+             EXPECT_CALL(handler, handle_input(mt::MotionEventWithPosition(80, 170))).Times(AnyNumber())
+                 .WillOnce(mt::WakeUp(&events_received));
+        });
+
+    launch_client_process(*client);
+}