aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* GLJPanel: Refine API docSven Gothel2013-03-261-8/+15
|
* Android/NEWT MouseEvent: Fix Detection/Processing of 2-Finger-Scroll Gesture ↵Sven Gothel2013-03-244-88/+154
| | | | | | | | | | | | | | | | | - Part 2 Multiple tests on different devices disclosed that: - 1 of 2 pointers get disconnected every now and then .. -> Shall tolerate this case - dist-delta within gesture may shrink below doubleTouchSlope -> Remove constraint after gesture detection - Always validate pointer-id GearsES2 - Works quite stable on several devices now - Moved soft-keyboad show to 4-pointer pressed >= 0.7f pressure
* Newt/MouseEvent: Add 'float[3] getRotation()', getWheelScale() -> ↵Sven Gothel2013-03-244-51/+107
| | | | | | | | | | | | getRotationScale(), refinement of commit 18cb57246372eda72c25a5cd9a69a873bdf09489 Turns out the 'wheel' semantics are not generic enough and confining rotation values to one axis only satisfies the traditional mouse wheel. Widen the definition of 'rotation' and delivering 3-axis if supported. On NEWT/Android, we deliver the 2-axis for example, allowing to rotate around both or scrolling using both directions (-> GearsES2).
* Android/NEWT MouseEvent: Fix Delivery of MultiTouch PRESSED/RELEASE; Fix ↵Sven Gothel2013-03-242-172/+305
| | | | | | | | | | | | | | | | | | | | | | | | Detection/Processing of 2-Finger-Scroll Gesture ; GearsES2: Add NEWT based 'zoom' gesture detection. - Fix Delivery of MultiTouch PRESSED/RELEASE Adopting MouseEvent changes of commit 18cb57246372eda72c25a5cd9a69a873bdf09489 - Fix Detection/Processing of 2-Finger-Scroll Gesture Dropping utilization of Android's GestureDetector and implementing our own, which turns out to simplify keeping track of states. Our gesture detection works well w/ user NEWT based gesture detection (-> See GearsES2 zoom and rotate), using following criteria related to Android parameter: - ScaledDoubleTapSlop: - Max 2 finger distance - ScaledTouchSlop: - Min. movement w/ 2 pointer withing ScaledDoubleTapSlop starting 'scroll' mode - Max. change of finger distance in respect to initiating 2-finger distance (2x ScaledTouchSlop) - Max. distance growth in respect to initiating 2-finger distance. - GearsES2: Add NEWT based 'zoom' gesture detection.
* NEWT/MouseEvent: Refine MultiTouch Event Spec regarding associated 'action' ↵Sven Gothel2013-03-241-15/+82
| | | | | | | | | | | | | | | | | | | | pointer Index; Add PointerType[PointerClass]; Add 'wheelScale' Attribute; - Refine MultiTouch Event Spec regarding associated 'action' pointer Index In case an instance represents multi-touch events, i.e. {@link #getPointerCount()} is > 1, the first data element represents the pointer which triggered the action if individual to one pointer. For example {@link #getX(int) e.getX(0)} at {@link #EVENT_MOUSE_PRESSED} returns the data of the pressed pointer, etc. - Add PointerType[PointerClass] This allows applications to identify the type and it's class [On-/Offscreen] helping to interpret semantics, e.g. wheel-rotate and new wheel-scale - Add 'wheelScale' Attribute Returns the scale used to determine the {@link #getWheelRotation() wheel rotation}, which semantics depends on the {@link #getPointerType() pointer type's} {@link PointerClass}. See API doc for details .. TODO: NEWT/Android changes adopting these changes.
* NEWT/Android: Fix suspense regression of commit ↵Sven Gothel2013-03-231-2/+6
| | | | | | | | | | d514ecbf052d013ea8c0982c490757678075a9ea Explicit win.destroy() shall only be called at activity.onDestroy(), - GLStateKeeper preservation is marked at pause, - and the Window's surfaceDestroyed() will also issue destroy() - so it's safe Reason: On Suspense (or power button), application is paused _without_ surfaceDestruction !
* NEWT/Android (Bug 665): Add Support for GLStateKeeper ; onPause() always ↵Sven Gothel2013-03-2216-76/+188
| | | | | | | | | | | | | | | | | | | destroys ; Recognizing all GLAutoDrawable's GLAnimatorControl for pause/resume ; Use GLAnimatorControl instead of Animator - Add Support for GLStateKeeper If !isFinishing() (HOME button), preserve the GLEventListener if an GLStateKeeper instance - onPause() always destroys onDestroy() is too late, i.e. surfaceDestroyed() already called - Recognizing all GLAutoDrawable's GLAnimatorControl for pause/resume pause/resume the GLAnimatorControl of all GLAutoDrawable instances - Use GLAnimatorControl instead of Animator We used an Animator reference .. duh! Note: The EGL native WindowDriver (Android and BCM.IV) must retain their own copy of EGLGraphicsDevice, which preserves the EGLDisplay handle due to EGLDisplayUtil reference counting per nativeHandleID.
* NEWT/Android WindowDriver.closeNative0(..): Catch Exception at surface ↵Sven Gothel2013-03-221-4/+10
| | | | destroy allowing to continue destruction.
* Animator: Cleanup DEBUG outputSven Gothel2013-03-221-4/+4
|
* Add GLStateKeeper handling GLEventListenerState preservation and ↵Sven Gothel2013-03-2211-24/+152
| | | | | | | | | | | | | | | restauration for GLAutoDrawable - New GLStateKeeper interface, package com.jogamp.opengl Implemented by: - GLAutoDrawableBase Currently supported by: - NEWT GLWindow - GLEventListenerState package move: com.jogamp.opengl.util -> com.jogamp.opengl
* NEWT/Android: Remove KeyEvent.KEYCODE_HOME handling, since it doesn't work - ↵Sven Gothel2013-03-221-19/+0
| | | | i.e. cannot be intercepted this way.
* NEWT/Android: Fix BACK button implementation, use different KeyCode mappings ↵Sven Gothel2013-03-215-24/+146
| | | | | | | | | | | | | | | | | | | | | and allowing native action to be suppressed. - Don't trust soft-kbd visibility state, but perform invisible action. If the latter was successful - soft-kbd was visible before. - Map BACK to VK_KEYBOARD_INVISIBLE and propagate it, if soft-kbd was visible before. No native default action is performed. - Map BACK to VK_ESCAPE event and propagate it, if soft-kbd was invisible _and_ an activity was registered via registerActivity(Activity), i.e. by NewtBaseActivity. Otherwise proceed w/ default action (-> activity.finish()). - If the KeyListener consumed the [EVENT_KEY_RELEASED, VK_ESCAPE] event, it will be suppressed and no default action performed. This allows applications to have a custom 'ESCAPE' or 'BACK' handling. Otherwise (not consumed) the default action is performed.
* NEWTEvent: Add isConsumed() and setConsumed(boolean) methods to simply ↵Sven Gothel2013-03-213-14/+42
| | | | usage, using the existing consumedTag attachment for compatibility and efficiency.
* OSX/NEWT: Fix visible -> false when closing and reparenting window. This ↵Sven Gothel2013-03-203-6/+4
| | | | | | | | | | | lead to non functional recreational reparenting. OSX recreational reparenting moves the saved GLEventListenerState at destroy to the new dawable/surface, which must be valid. The flaky visible state caused seemingly random reparenting failures. - WindowImpl.ReparentActionRecreate.run() set 'visible:=true', which circumvented OSX to wait for actual realization. - OSX WindowDriver.closeNative(): Issue visibleChanged(true, false); ASAP
* MacOSXCGLContext: NPE in DEBUG mode while releaseNSOpenGLLayer on ↵Sven Gothel2013-03-201-2/+2
| | | | main-thread, drawable maybe null already
* OSX/NEWT: Fix native window parenting freeze, invisible/orderOut0, Position ↵Sven Gothel2013-03-204-75/+101
| | | | | | | | | | | | | | | | | | | | | | | | | | | | bugs w/ parenting Fix native window parenting freeze: - Pull out setJavaWindowObject(..) of changeContentView(..) to be called seperately, add param for changeContentView(..) to enable/disable setJavaWindowObject() - initWindow0(..): - Call changeContentView(..) w/o setJavaWindowObject() - setJavaWindowObject(..) at end of initialization Fix native window parenting orderOut0: If parent window is invisible or no parent used call orderOut(..), otherwise call orderBack(). Fix updatePosition(..): positionChanged(..) - Position bug w/ parenting - AWT parent passed 0/0 - call positionChanged(..) w/ client-pos instead of screen-pos Fix getLocationOnScreenImpl(..) - Position bug w/ parenting - Position < 0/0 is valid! Misc: - setWindowClientTopLeftPointAndSize0(..), setWindowClientTopLeftPoint0(..): Add 'display' param, deciding whether area should be display (invalidated)
* TestGearsES2NEWT/TestGearsES2NewtCanvasAWT: Add -noanim option to manually ↵Sven Gothel2013-03-203-31/+64
| | | | validate Bug 649
* PNGImage: Cleanup code and reuse scanlineRGBA buffer in indexed mode. Add ↵Sven Gothel2013-03-201-29/+36
| | | | DEBUG property 'jogl.debug.PNGImage'
* TestPNGTextureFromFileNEWT: Use new PNG test file locations, add RGBA tests ↵Sven Gothel2013-03-202-83/+89
| | | | and use alpha in drawable
* Newt PNG test file locationsSven Gothel2013-03-202-2/+2
|
* Rename PNG test files and add RGBA variations (To be used by ↵Sven Gothel2013-03-209-0/+0
| | | | TestPNGTextureFromFileNEWT, etc)
* Handles indexed PNGsJulien Gouesse2013-03-204-21/+117
|
* OSX/NEWT: NSWindow/NSView Ops on main-thread w/o blocking - Part2Sven Gothel2013-03-192-62/+117
| | | | | | | | | Continues commit 81cbcdc8469143587b2044661dd613c798ae02ba Perform on main-thread invocation from Java, allowing to issue visibleChanged(..) after creation/visible calls. This fixes the 'Visibility not reached ..' regressions.
* MacOSXCGLContext: Remove Debug outputSven Gothel2013-03-191-1/+0
|
* GLJPanel: No need to utilize AWT Component's default ImageObserver code - ↵Sven Gothel2013-03-191-1/+1
| | | | pixel data is ready when invoked.
* OSX/NEWT: Following CALayer streaming design, i.e. issue NSWindow/NSView Ops ↵Sven Gothel2013-03-197-233/+376
| | | | | | | | | | | | | | | | | | | | | on main-thread w/o blocking; NEWT/WindowImpl: Volatile multithreaded mutable values Similar to commits: 28c6472335b924080d638b33a28f8f4eedb459b1 f354fb204d8973453c538dda78a2c82c87be61dc main-thread operations cannot block main-thread. Luckily we are able to create the NSWindow and NSView instance uninitialized (deferred) on the current thread, while issuing their initialization on the main-thread w/o blocking. Further more a size glitch is fixed, which didn't take the title bar into account. +++ NEWT/WindowImpl: Volatile multithreaded mutable values Since position, size and other attributes might get changes off-thread, these fields needs to be volatile.
* OSX/CALayer: OSX/CALayer Threading Part3 - Run CALayer ops in a streaming ↵Sven Gothel2013-03-1914-259/+439
| | | | | | | | | | | | | | | | | | | | | | | | | | | design on main-thread w/o [infinitive] blocking History: Part1 commit 896e8b021b39e9415040a57a1d540d7d24b02db1 (Run CALayer Ops on current thread to avoid blocking) Part2 commit 28c6472335b924080d638b33a28f8f4eedb459b1 (Run CALayer Ops on main-thread w/o blocking) Dependency: GlueGen commit 4becdfa125b07ff969d6540e1112735b53cd15eb (Fix RecursiveLockImpl* Timeout corner case) Part2 misses essential locking of the OpenGL context (and it's surface upfront) while creating the NSOpenGLLayer instance. The latter instantiates a OpenGL context shared w/ JOGL's, hence it cannot be locked. Encapsulating NSOpenGLLayer creation/attachment and it's detachment/release in sub-classes AttachNSOpenGLLayer and DetachNSOpenGLLayer, where instances will be streamed on main-thread. Both tasks are triggered at associateDrawable(boolean bound). The mentioned GL context locking requires disturbs the 'streaming' design considerably in AttachNSOpenGLLayer. It is solved by attempt to acquire the recursive lock of the surface and the context via 'tryLock(maxwait)' w/ screen-vSync-period/2. If the locks could not be acquired completly, the AttachNSOpenGLLayer instance will be re-queued to the main-thread for later execution. Before DetachNSOpenGLLayer is being streamed, it is validated whether AttachNSOpenGLLayer did run. A recursive situation does happen w/ resizing an offscreen pbuffer drawable! Hence extra care is being taken.
* MyNSOpenGLContext::dealloc: Avoid 'invalid context'Sven Gothel2013-03-192-7/+29
|
* OSX CGL: Don't issue [NSOpenGLContext clearDrawable] for [NSOpenGLContext ↵Sven Gothel2013-03-182-3/+11
| | | | setView: view] which breaks pbuffer; Add [NSOpenGLContext clearDrawable].
* NativeWindow OSXUtil RunOnMainThread: Use daemon attachment and do not ↵Sven Gothel2013-03-184-27/+53
| | | | detach; Add RunLater0(..)
* Adapt to GlueGen commit b1eb7ca6b9d7dec7ff62c1f1e8ef0a0545724d2f: Function- ↵Sven Gothel2013-03-186-8/+10
| | | | RunnableTask adds PrintStream 'exceptionOut' argument in ctor.
* Remodel OSX/CALayer Threading (commit ↵Sven Gothel2013-03-1515-249/+358
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 896e8b021b39e9415040a57a1d540d7d24b02db1): Run on main-thread w/o blocking ; Misc Changes Commit 896e8b021b39e9415040a57a1d540d7d24b02db1 moved all native CALayer calls to the current thread to avoid deadlocks. Even though this seemed to be fine at least resource GC (release/dealloc calls) were issued very late in time, probably due to multithreading synchronization of JAWT and/or OSX API. Example: Our 'TestAddRemove01GLCanvasSwingAWT' test didn't freed CALayer resources incl. GL ctx when destroying the objects (AWT Frame, GLCanvas, ..), leading to resource starvation .. eventually. Remedy is a compromise of behavior before commit 896e8b021b39e9415040a57a1d540d7d24b02db1 and that commit, i.e. to run CALayer lifecycle methods on main-thread, but do not block! The careful part within MacOSXCGLContext.associateDrawable(..) performs the following block on main-thread: - lock the context - create NSOpenGLLayer (incl. it's own shared GL context and the DisplayLink) - attach NSOpenGLLayer to root CALayer - unlock the context Due to the GL ctx locking, this async offthread operation is safe within our course of operations. Details: - NSOpenGLContext - Context and CVDisplayLink creation at init - Call [ctx update] if texture/frame size changed - 'waitUntilRenderSignal' uses default TO value if given TO is 0 to avoid deadlocks +++ Misc Changes: - Fix object type detection: isMemberOfClass -> isKindOfClass - OSXUtil_isNSView0 OSXUtil_isNSWindow0, CGL_isNSOpenGLPixelBuffer - MacOSXCGLDrawable/MacOSXPbufferCGLDrawable: remove getNSViewHandle() method. MacOSXCGLContext uses common code to detect nature of the drawable handle. - MacOSXCGLContext/CALayer: Use safe screenVSyncTimeout values, never 0 to avoid deadlock! - JAWTWindow.invalidate: Call detachSurfaceLayer() if not done yet
* Fix NEWT WindowImpl reparent-recreate w/ GLEventListenerState: Bug ↵Sven Gothel2013-03-143-17/+21
| | | | | | | | | introduced w/ commit e2506d7663b752f00f0a98f793ebad52e65bd1e3 In case a reparent action takes place w/ recreate, only preserve the GLEventListenerState if the window is valid and will become visible again (wasVisible). Also add proper DEBUG log prefix to GLEventListenerState.
* OSX/CALayer: Simplify FixCALayerLayout()/layoutSurfaceLayer() call, no more ↵Sven Gothel2013-03-147-94/+46
| | | | | | | | | | | | need for explicit call - OffscreenLayerSurface.layoutSurfaceLayer() removed, no more required - JAWTWindow adds a ComponentListener, which issues FixCALayerLayout() at resized, moved and shown. - MyNSOpenGLLayer no more requires fix*Size() methods - MyNSOpenGLLayer::setDedicatedSize() need no explicit CATransaction, performed by caller.
* OSX/CALayer: Revise CALayer 'RunOnMainThread' utilization, avoiding deadlocksSven Gothel2013-03-148-123/+102
| | | | | | | | | | | | | | | | | | | | | | RunOnMainThread(waitUntilDone:=true,..) can deadlock the main-thread if called from AWT-EDT, since the main-thread may call back to AWT-EDT while injecting a new main-thread task. This patch revises all RunOnMainThread CALayer usage, resulting in only one required left: - OSXUtil.AddCASublayer() w/ waitUntilDone:=false Hence the CALayer code has no more potential to deadlock main-thread/AWT-EDT. OSXUtil.AddCASublayer() must be performed on main-thread, otherwise the CALayer attachment will fail - no visible rendering result. +++ Note: A good trigger to test this deadlock is to magnify/zoom the OSX desktop (click background + ctrl-mouse_wheel) before running some unit tests. TestGLCanvasAWTActionDeadlock01AWT and TestAddRemove02GLWindowNewtCanvasAWT also have the potential to trigger the mentioned deadlock.
* Fix TestParenting01NEWT: Since recreational reparenting resets the ↵Sven Gothel2013-03-131-0/+6
| | | | fps-counter, wait for a few frames!
* GLEventListenerState: Moved to public package 'com.jogamp.opengl.util'Sven Gothel2013-03-138-13/+14
|
* Minor Changes: EGLDisplayUtil: Add stack trace in DEBUG mode for opened ↵Sven Gothel2013-03-134-30/+85
| | | | DPYs; JAWTWindow: Add JAWT info in toString()
* Bug 665: Allow re-association of GLContext/GLEventListener to a GLDrawable ↵Sven Gothel2013-03-133-17/+89
| | | | | | | | | | | | | | | | | | | | (Part 5) - GLAutoDrawableBase: - Add 'setPreserveGLStateAtDestroy(..)' to preserve the GLEventListenerState at destroy() operation, and impl. details pullGLEventListenerState()/pushGLEventListenerState(). pullGLEventListenerState() is called automatic at destroyImplInLock(), where pushGLEventListenerState() has to be called after drawable realization instead of context creation. - Note/TODO: Method will become public in GLAutoDrawable in general! - NEWT/GLWindow: - Use GLEventListenerState preservation for reparenting case w/ destruction, i.e. keep GLContext/GLEventListener alive while reparenting in recreation mode. Scenario: NewtCanvasAWT Child <-> Top on OSX w/ CALayer
* Bug 665: Allow re-association of GLContext/GLEventListener to a GLDrawable ↵Sven Gothel2013-03-1326-442/+1113
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (Part 4) Note: - GLEventListenerState preservs the GLAutoDrawable state, i.e. GLContext, all GLEventListener and the GLAnimatorControl association. - GLEventListenerState may be utilized to move the state from a dying GLAutoDrawable, to be moved to a new created GLAutoDrawable at a later time. - GLEventListenerState will be made public soon. +++ Exessive unit tests cover the new feature, tested manually on GNU/Linux/X11 and OSX(Java6/Java7). +++ - GLAutoDrawable - Change 'setContext(..)' to allow the destruction of the unbound old context: 'setContext(GLContext newCtx)' -> 'setContext(GLContext newCtx, boolean destroyPrevCtx)' - Implementations: Properly implement 'setRealized(..)' incl. obeying threading constraints if exists. Method is being utilized at least for GLEventListenerState.moveTo(..) to unrealize and realize the drawable resources. +++ Fix propagation of GLContext/GLDrawable association change (Bottom -> Top): GLDrawableImpl.associateContext GLContextImpl.associateDrawable GLContextImpl.makeCurrent GLContextImpl.destroy GLContext.setGLDrawable ... GLDrawableHelper.switchContext GLAutoDrawble.setContext associateDrawable(..)/associateContext(..) unifies and hence: - GLContextImpl.contextRealized() (removed) - GLDrawableImpl.contextRealized() (removed) - GLDrawableImpl.associateContext(..) (merged) - MacOSXCGLContext.drawableChangedNotify(..) (removed) +++ - EGLUpstreamSurfaceHook.evalUpstreamSurface() validates the surface's device for reusage, which is valid in case of GLEventListenerState.moveTo(..) - MacOSXCGLContext.NSOpenGLImpl: pixelFormat replaces NSOpenGLLayerPfmt and has simplified lifecycle [create..destroy], while native NSOpenGLLayer code only holds the reference until released.
* MemoryObject: Use GlueGen's new HashUtilSven Gothel2013-03-131-44/+6
| | | | GlueGen commit 1a4514accc8f61ab7ff5fe8c82d22a5ef356c865
* *External*Context Impl: Remove unused 'lastContext' field, incl. their ↵Sven Gothel2013-03-134-85/+2
| | | | | | | makeCurrent() and release() override. Semantics of 'lastContext' are plain wrong, since release() override does claim the previous 'lastContext' is current at the end - however, it wasn't technically made current.
* Fix Mix2TexturesES2 unit tests: Use synchronized texID access and add ↵Sven Gothel2013-03-095-18/+90
| | | | | | glFinish() after demo -> FBO rendering on or off thread. Fixes shared GLContext OSX rendering.
* Fix buggy unit test for Bug 694: The unpack alignment has to be considered!Sven Gothel2013-03-092-18/+74
|
* GLJPanel: Updating GLJPanel documention w/ current mode of operationSven Gothel2013-03-081-12/+14
|
* GLJPanel: Add DEBUG_VIEWPORT flag, Add USE_GLSL_TEXTURE_RASTERIZER flag, ↵Sven Gothel2013-03-081-25/+42
| | | | | | | | | | | | | | DEBUG reshape trace. - Add DEBUG_VIEWPORT flag Via explicit property 'jogl.debug.GLJPanel.Viewport' (not via jogl.debug=all) Traces the current GL Viewport in OffscreenBackend.postGL(..) - Add USE_GLSL_TEXTURE_RASTERIZER flag Via explicit property 'jogl.gljpanel.noglsl'. Disables use of GLSL FBO flipping. DEBUG reshape trace
* TestBug694 -> TestBug694AWT so it properly gets picked up by the AWT unit ↵Sven Gothel2013-03-011-2/+2
| | | | test ant task.
* Merge remote-tracking branch 'gouessej/master'Sven Gothel2013-03-011-0/+128
|\
| * Adds a unit test for the bug 694 on gluScaleImageJulien Gouesse2013-02-241-0/+128
| |
* | Fix Bug 695: WGLExt.wglChoosePixelFormatARB causes buffer underflow due to a ↵Sven Gothel2013-02-282-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | higher reported number of configs than buffer size "I encountered a case on an NVidia Quadro 3500 fx where the call to WGLExt.wglChoosePixelFormatARB in WindowsWGLGraphicsConfiguration (currently line 355) returns 264 in numFormatsTmp despite 256 being passed in for the maximum number of formats. This results in a buffer underflow on line 368 since pformatsTmp only has 256 values and it's trying to copy 264 values." Fixed in WindowsWGLGraphicsConfiguration and WindowsPbufferWGLDrawable.