Changes

GLContextImpl: Move sharedContextHandle check to makeCurrentWithinLock(..) and let it fail there instead of within impl. class, only pass the handle - simplifies and removes redundancy.

Bug 888 - Validate CPU Runtime Performance: X11GLXGraphicsConfiguration.GLXFBConfig2GLCapabilities(..)

X11GLXGraphicsConfiguration.GLXFBConfig2GLCapabilities(..) ran over all FB configs and for each it grabbed
native config values separately. Fetching them in bulk mode saves around 7% of this function's cost.

Also reuse XRenderPictFormat instance for 'XRenderDirectFormat XRenderFindVisualFormat(..)' call,
saving a few NIO creation cycles w/ StructAccessor.

Biggest savior is X11GLXGraphicsConfigurationFactory.chooseGraphicsConfigurationFBConfig()'s
fast path w/o chooser and usable 1st FBConfig. Here we only issue 'GLXFBConfig2GLCapabilities(..)'
on the first valid entry.

Test w/ 50 X11GLXGraphicsConfigurationFactory.chooseGraphicsConfigurationFBConfig() invocations:
  - pre change:  1.708 ms
  - post change:   650 ms

Time is no spent almost solely on native glXChooseFBConfig (546ms).

Bug 888 / Bug 891 - Enhance GLCapabilities-Query: Apply changes of commit 613e33ee8ffc1f2b9c5db1e1b5bb5253a159ed6d to EGL and WGL.

Note: WGL config query is already performed as a bulk operation.
Note: OSX does not perform such queries.

Bug 888 - Validate CPU Runtime Performance: Only use one HashMap to collect unified gl- and glx-extension strings.

- HashMap is more efficient than HashSet
- No need to use sub HashSet's .. we can use global HashMap.
  'Loosing' duplicate GLX entries due to GL duplicates is acceptable.

Test DumpGLInfo: Dump caps list and extenions

Bug 888 / Bug 891- Enhance GLCapabilities-Query: Add perf. test case

Even though the test case itself cannot show the proper initialization time,
it can be used w/ an attached profiler i.e.

    Test w/ 50 X11GLXGraphicsConfigurationFactory.chooseGraphicsConfigurationFBConfig() invocations:
      - pre change:  1.708 ms
      - post change:   650 ms  (613e33ee8ffc1f2b9c5db1e1b5bb5253a159ed6d)

Bug 888 / Bug 891- Enhance GLCapabilities-Query: Add raw perf. test case 'choose' only and 'full' offscreen-drawable w/ context

test01ChooseOnly exposes X11/GLX perf. enhancement of 613e33ee8ffc1f2b9c5db1e1b5bb5253a159ed6d:

PRE CHANGE:
++++ UITestCase.setUp: com.jogamp.opengl.test.junit.jogl.perf.TestPerf001RawInit00NEWT - test01ChooseOnly
INIT START #0
Run: 0, count 50/50 raw:
        choose     503/t  10.06/1
INIT END #0
INIT START #1
Run: 1, count 50/50 raw:
        choose     384/t   7.68/1
INIT END #1
INIT START #2
Run: 2, count 50/50 raw:
        choose     344/t   6.88/1
INIT END #2

POST CHANGE:
++++ UITestCase.setUp: com.jogamp.opengl.test.junit.jogl.perf.TestPerf001RawInit00NEWT - test01ChooseOnly
INIT START #0
Run: 0, count 50/50 raw:
        choose      49/t   0.98/1
INIT END #0
INIT START #1
Run: 1, count 50/50 raw:
        choose      43/t   0.86/1
INIT END #1
INIT START #2
Run: 2, count 50/50 raw:
        choose      38/t   0.76/1
INIT END #2

JNI Code: Call DeleteLocalRef(..) manually.

Bug 888 / Bug 891- Enhance GLCapabilities-Query: Refine unit tests

NEWT Window: Expose 'setVisible(boolean wait, boolean visible)' allowing applications to not block until window becomes visible.

NativeWindow *GraphicsDevice: Align constructors to simplify call hierarchy for analysis.

Bug 894 - GLDrawableFactory* [dummy|offscreen] Surface creation w/ own device does _not_ require locking on global shared device.

Bug 894 - GLJPanel: Expose 'initializeBackend(boolean offthread)' allowing user to trigger backend initialization eagerly and offthread (optional, !WINDOWS)

TestPerf001GLJPanelInit02AWT compares all variations: no-gl, glcanvas, gljpanel and gljpanel-initMT (offthread)

TestPerf001GLWindowInit03NEWT: Don't run !reuse Display - Crash on Windows/ATI driver ..

Android: Add NewtDebugActivity

Android: Cleanup Movie*ActivityLauncher* properties

Android NEWT*Activity: Request GL2ES2 instead of GLES2 (Be aware of ES3)

Bug 890 - Fix GLES3 Profile Mapping, i.e. GL2ES2 queries and mappings; Validate isGLES*() usage and definition ; Add and use ShaderCode.createExtensionDirective(..)

- Fix GLES3 Profile Mapping, i.e. GL2ES2 queries and mappings
  - GLProfile: Add GL2ES2 -> ES3 mapping
  - EGLContext: Reuqest major '3' for ES3
  - EGLGLCapabilities/EGLGraphicsConfiguration: Consider EGLExt.EGL_OPENGL_ES3_BIT_KHR

- Validate isGLES*() usage and definition
  - Fix BuildComposablePipeline's isGLES() code

  - For GLSL related queries use isGLES() instead of isGLES2(),
    which would exclude ES3

- Add and use ShaderCode.createExtensionDirective(..)
  - Supporting creating GLSL extension directives while reusing strings from GLExtensions

- Minor cleanup of GLContextImpl.setGLFuncAvail(..)

Android Newt[Debug|Version]Activity: Separate both activities (testing odd reusing of app)

Bug 898: Apply best efford on Animator operation's finishLifecycleAction(..) in !blocking mode, i.e. notifyAll() if waitCondition holds and test again

GLJPanel: Remove redundant !isInitialized [double-]check in initializeBackendImpl(), already tested before function entry!

TestTiledPrintingGearsSwingAWT: Cleanup whitespaces

WGLGLCapabilities: Print pfdID as decimal (convention)

WindowsWGLGraphicsConfigurationFactory: Fix chosenPFDID -> recommendedIndex (!skipCapsChooser mode)

Regression of commit cf1163fc88976e7087d3a17524a49139e35a4708:
  Commit dropped seeking recommendedIndex of chosenPFDID within cleaned-up availableCaps
  when in !skipCapsChooser mode.

GLJPanel: Don't ctor DefaultGLCapabilitiesChooser() if null chooser is passed, this allows native GLCaps/config query to follow fast-path

Android ES3 Movie Demos: Add workaround for: P0003: Extension 'GL_OES_EGL_image_external' not supported

+                // Bug on Nexus 10, ES3 - Android 4.3, where
+                // GL_OES_EGL_image_external extension directive leads to a failure _with_ '#version 300 es' !
+                //   P0003: Extension 'GL_OES_EGL_image_external' not supported
+                preludeGLSLVersion = false;

TestPerf001GLJPanelInit02AWT: Add overlapping test ..

TestPerf001GLJPanelInit02AWT: Same size for !overlap and overlap

Tests: Add com/jogamp/opengl/test/junit/jogl/acore/anim Animator test package

Simplify Animator* Synchronization: Remove barrier 'stateSync' and favor simple 'synchronized' on Animator for field-get, which is already used in most methods

Utilizing a 2nd synchronization object 'stateSync' besides the main sync object, Animator itself,
is hard to maintain. It's performance advantages for querying states ae questionable and may even
introduce bugs.

Use synchronization on Animator instance for all field read/write access.

Fix unsynchronized write access of 'animThread' in Animator.MainLoop.run().

Tests: Align NEWT parenting focus tests (package and class names) using NewtCanvasAWT

Fix Bug 901 - NEWT: Lookup Windows 7 Touch Event Functions Dynamically to not break Windows < 7 Compatibility

Bug 899: Validate whether we propagate WNDPROC as retrieved from 'getDummyWndProc0()' in RegisteredClassFactory - Result: Yes we do.

GDI.initSingleton() dummyWindowClassFactory RegisteredClassFactory[moduleHandle 0x13f3e0000, _dummyWindow_clazz, wndProc 0x6c101de6, shared[refCount 0, class null]]

GDI.CreateDummyWindow() dummyWindowClassFactory RegisteredClassFactory[moduleHandle 0x13f3e0000, _dummyWindow_clazz, wndProc 0x6c101de6, shared[refCount 1, class RegisteredClass[handle 0x13f3e0000, _dummyWindow_clazz0]]]
GDI.CreateDummyWindow() dummyWindowClass RegisteredClass[handle 0x13f3e0000, _dummyWindow_clazz0]

++

Note: The RegisteredClassFactory mechanism is used for NEWT Windows as well.

Add TestGLWindows03NEWTAnimResize: Test NEWT basic resize while holding position .. (to be refined)

Fix Bug 889 [Related: Bug 816, Bug 849, Bug 729] - GLCanvas disappear when moves between two JFrame

When JAWTWindow's visibility tracker updates component's local visibility,
it should read it's local visibility state instead 'trusting' the passed state.

Make JAWTWindow's visibility tracker DEBUG output more brief for readability.

Fix javadoc: TileRendererBase (links to manual anchors) and GLSharedContextSetter (remove <> in links, more details on Immersion.16.

GLJPanel: Remove unused import

NEWT: Add more documentation to WindowImpl's doPointerEvent(..) and consumePointerEvent(..) impl. details

Bug 904 - GLJPanel: Allow user to skip isGLOriented() based vertical flip of offscreen backend

Add new GLJPanel method 'setSkipGLOrientationVerticalFlip(..)':

  /**
   * Set skipping {@link #isGLOriented()} based vertical flip,
   * which usually is required by the offscreen backend,
   * see details about <a href="#verticalFlip">vertical flip</a>
   * and <a href="#fboGLSLVerticalFlip">FBO / GLSL vertical flip</a>.
   * <p>
   * If set to <code>true</code>, user needs to flip the OpenGL rendered scene
   * <i>if {@link #isGLOriented()} == true</i>, e.g. via the PMV matrix.<br/>
   * See constraints of {@link #isGLOriented()}.
   * </p>
   */
  public final void setSkipGLOrientationVerticalFlip(boolean v) {

GearsES2: Handles 'flipVerticalInGLOrientation'

Unit test 'TestPerf001GLJPanelInit02AWT' validates and measures performance.

Bug 904 - GLJPanel: Add property to skip isGLOriented() based vertical flip by default (2/2)

Property 'jogl.gljpanel.noverticalflip' will set the skipGLOrientationVerticalFlip default to true - intended for perf. testing of existing applications

TestPerf001GLJPanelInit02AWT: Wrong test name 'Gears' -> 'Nop'

Applet Test Bug848AppletGLCanvas01: Only use vsync-interval 1 for last GLJPanel

Bug 906 - JAWTWindow Component- and HierarchyListener must be detached at JAWTWindow.destroy() - GLCanvas Recreation Case

In case a JAWTWindow owner recreates itself, destroying JAWTWindow must detach the
Component- and HierarchyListener from the AWT component!

Test TestBug816GLCanvasFrameHoppingB849B889AWT
shows properly detaching listener at jawtWindow destruction.

Bug 903 - NEWT: Support 'Continue Drag on Exit'; Consistent Mouse ENTER/EXIT

- Support 'Continue Drag on Exit'
  - Track dragging operation, allow exterior dragging
  - Hence track EXIT (see below)
  - Windows:
    - Capture mouse for exterior dragging
    - Only 'NewtWindows_trackPointerLeave' if 'entering'
    - Simplify touch: No 'inside' check - Not required.

- Consistent Mouse ENTER/EXIT
  - Track ENTER/EXIT and synthesize if required, drop duplicate
  - OSX benefits, since it never produced ENTER/EXIT events
  - AWT (or other TK) translated events beahve equal now.
  - Required for EXIT event after ending exterior dragging and final RELEASE

Tests: Passed unit tests 'junit.run.newt.event' on
  - GNU/Linux
  - Windows7
  - OSX 10.7

Tested exterior tracking manually w/ NEWT TestGearsES2NEWT and TestGearsES2NewtCanvasAWT:
  - GNU/Linux
  - Windows7 (mouse)
  - Windows8.1 (touch)
  - OSX 10.7

Fix Bug 879 - Threads deadlock in native keyboardfocus calls made form multiple threads; Fix Bug 892: Reduce Focus Hopping

Since we manage focus key traversal ourselves w/o requiring the AWT
component to have the focus[1],
we simply can drop requesting the focus for 'focus hopping' NEWT -> AWT -> NEWT[2].

Further more, 'MenuSelectionManager.defaultManager().clearSelectedPath()'
must be performed on AWT-EDT w/o blocking. Otherwise it may perform blocking tasks on AWT-EDT.

[1] Commit cb7118fc875b6722803e4b11d5681671962a8d3a
introduced function to query the next or previous 'to be focused' component:
   AWTMisc.getNextFocus(..) .. etc.

[2] Focus hopping is also addressed in Bug 892

Fix GLIBC > 2.4 dependency regression of commit 613e33ee8ffc1f2b9c5db1e1b5bb5253a159ed6d

Commit 613e33ee8ffc1f2b9c5db1e1b5bb5253a159ed6d introduced 'memcpy' usage in Xmisc.c which could create a GLIBC > 2.4 dependency.
Include GlueGen's glibc-compat-symbols.h to remove such dependency.

Nativewindow/NEWT: Fix C Return Statement

Use 'gluegen-clang.properties' for generic clang and 'gluegen-xcode_clang.properties' for OSX xcode-clang ; Add GNU/Linux LLVM/clang build scripts

Fix Bug 879 Regression (2/2) - NewtCanvasAWT.FocusAction must take focus when in offscreen-mode (OSX/CALayer)

NewtCanvasAWT.FocusAction must take focus when in offscreen-mode (OSX/CALayer)
since the NEWT window _is_ offscreen (no input events) and AWT events are translated to NEWT.

Regression of commit 0be87f241c0f0b2f5881d9a602ce12378b8e453d

NEWT AWTAdapter: Add notion of consuming the AWT InputEvent (will be used for key events); Allow AWTAdapter to be lazily setup w/ downstream object.

NewtCanvasAWT: Use final AWT[Key|Mouse]Adapter and set downstream lazily; Consume AWT KeyEvents in downstream mode; Test respects 'consumed' key events.

NEWT AWTAdapter*: Don't act if not setup (due to lazy setup mode), refines commit 5c6c11abf643013976ecbc0df463a923a1f52696

NewtCanvasAWT: Add method of 'isAWTEventPassThrough()', used in unit tests to fix event validation for offscreen mode (OSX/CALayer)

NEWT Event Tests: Move TestParentingFocus* to event test package

NEWT WindowImpl: Move consumePointerEvent(..) below doPointerEvent(..) to easy editing/review

EGLGraphicsConfiguration: Add deauled DEBUG output for failed EGL-Config Queries

Bug 890 - GLContextImpl's Version Validation: Only validate Integer based Version if requested context or string-version is >= 3.0

.. otherwise, spec doesn't require integer based version to work!

Fix Bug 893 - NewtCanvasAWT Lifecycle Race Condition (NPE on shutdown periodically)

As suggested: Employ synchronization on lifecycle actions _and_ perform destroyImpl(..) always on AWT-EDT to avoid a deadlock.