| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
signaling no extension is available.
|
|
|
|
| |
JVM/OpenGL-Driver issues
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
12 -> 7 in initialization.
Reducing the calls to resetProcAddressTable() 12 -> 7 in initialization, saves:
Linux/AMD: 600ms -> 300ms
Linux/NV: 161ms -> 112ms
OSX 10.7/NV: 522ms -> 397ms
Still some freezes on OSX 10.6.8/NV .. further analysis is going on.
|
|
|
|
| |
is called 1st, add tests. Add performance numbers of init/shutdown in tests.
|
|
|
|
| |
hook or not
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
'dead' thread (Applets)
GLDrawableFactory:
- clarify: public getWasSharedContextCreated(..) -> protected createSharedResource(..)
- add: getSharesResourceThread()
GLProfile:
- proper locking of initSingletion(..) path:
- Use RecursiveThreadGroupLock and add/remove GLDrawableFactory's sharesResourceThread
while creating it's the sharedResource.
This simplifies and fixes GLProfile's locking code.
- Fix and simplify initSingleton(boolean) API doc - mark it deprecated.
- Add initSingleton() for controlled initialization only, pairing w/ shutdown(..)
Remove initSingleton(boolean) calls in code and test!
+++
Fix SharedResourceRunner 'dead' thread (Applets)
In Applets, stopping an Applet makes the browser Java plugin interrupting
and killing all related threads, including our SharedResourceRunner thread.
- Validate whether the shared resource thread is alive
- Catch interruption in shared resource thread and assume it's a kill signal
- releaseSharedResource: clear devicesTried set
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Management
- Leave Platform, .. TempJarCache untouched.
- GLDrawableFactoryImpl*: Leave DynamicLibraryBundle(lib-binding) untouched,
for NativeLibrary, JNILibLoaderBase (JNI libs), .. consistency.
- SHARED_ONLY: shutdown shared GLDrawableFactoryImpl* resources and NativeWindowFactory
- COMPLETE: additionally shutdown GLContext*
Clear all cached GL/GLX proc-address and device/context mappings.
- Use new "GLProfile.shutdown(GLProfile.ShutdownType.SHARED_ONLY)" in Applets
- X11GLXDrawableFactory Shutdown: Uncomment close/destroy of shared resources.
- JAWTWindow.destroy(): Close the delegated device.
In case it's X11 this closes the exclusive opened X11 Display.
|
|
|
|
| |
(GLContextImpl)
|
| |
|
| |
|
|
|
|
| |
constructor/destroy()
|
|
|
|
|
|
| |
RegisteredClassFactory: Reference the factories itself instead of the RegisteredClass.
This enables the shutdown hook to clear the factories state,
which is required for proper recreation.
|
|
|
|
|
|
|
|
|
|
|
|
| |
More Robust GLProfile Initialization
- Catch GLException in GLDrawableFactory getWasSharedContextCreated(device) impl.,
which may fail (See comment on Firefox/Chorme EGL deployed library for Windows).
- If getWasSharedContextCreated(devide) fails, set respective factory availability to false,
ie. hasDesktopGLFactory, hasEGLFactory, ..
Add NativeWindowFactory Shutdown
- Currenly a dummy entry, may evolve. X11Util shutdown is issued by respective GLDrawableFactory
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
createImpl(): If successful must leave context current.
makeCurrentImpl(): Is only called if context is not just created,
hence the boolean parameter 'boolean newCreatedContext' is removed.
This clearifies and actually cleans up the native makeContextCurrent/releaseContext call pairs.
MacOSXCGLContext: CGL and NS impl. of native makeContextCurrent/releaseContext uses CGL locking
to provide a thread safety. This is recommended in OS X OpenGL documentation on
[shared context] multithreaded use cases.
Post creation code, as seen in some pbuffer cases is moved to overriden createImpl() methods.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
use (threading)
Bug 527 describes the situation where wglShareLists() fails (throws exception)
while the shared context is in use by another thread (some Animator).
It was reported by Jerome Jouvie.
The exception happens not everytime, but at least around 20% on manual tests
I have performed on the Windows platform.
The context in question are all JOGL's GL2, which natively where bound to an OpenGL 3.0 profile.
The WGL_ARB_create_context spec says http://www.opengl.org/registry/specs/ARB/wgl_create_context.txt:
+++
Future versions of OpenGL may only support being added to a share
group at context creation time. Specifying such a version of a
context as either the <hglrc1> or <hglrc2> arguments to
wglShareLists will fail. wglShareLists will return FALSE, and
GetLastError will return ERROR_INCOMPATIBLE_DEVICE_CONTEXTS_ARB.
+++
Hence the 1st patch was to remove 'wglShareLists()' usage in case
we use the new WGL_ARB_create_context context creation method.
Even though this is a desired change, and works in general, it didn't fix the issue.
It seems that the shared context, which is passed @ new context creation,
cannot be used while it is in use itself in another thread.
This conclusion leads to the actual fix, ie. locking the shared context
while creating the new context which shares it.
Manual tests using this patch could not reproduce this issue (40 attempts).
Test: TestSharedContextListNEWT2
|
|
|
|
|
| |
We shall lock a level above due to differentiation of XLockDisplay/RecursiveLock
using X11/AWT ToolkitLock.
|
|
|
|
| |
shutdown: don't attempt to close Display device.
|
| |
|
|
|
|
|
|
|
| |
- GLXUtil: Distinguish between client and server GLX information, cache client information.
- GLXDrawableFactory: Utilize GLXUtil client data, as well as cache (SharedResource) GLX server data,
avoiding 'uncontrolled' GLX queries, ie. w/o locking.
- isMultisampleAvailable = isClientMultisampleAvailable && isServerMultisampleAvailable
|
|
|
|
| |
constructor
|
|
|
|
| |
AWTDevice: Remove subtype
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NativeWindow X11AWTGraphicsConfigurationFactory
X11AWTGraphicsConfigurationFactory properly construct a AWTGraphicsConfiguration
encapsulated a native X11GraphicsConfiguration, now available for non JOGL modules, ie NEWT.
AWTGraphicsConfiguration's create() utilizes the X11AWTGraphicsConfigurationFactory
via the generic factory mechanism and hence allows encapsulating a native [X11]GraphicsConfiguration.
NewtCanvasAWT creates/destroys the JAWT NativeWindow on addNotify/removeNotify (reparentWindow) again.
Hence the JAWTWindow is instantiated completly only, instead of utilizing updateConfiguration(..),
which simplifies the mechanism.
|
|
|
|
|
|
| |
instead of constructor for clarity.
- prepare for 'jogamp.nativewindow.x11.awt.X11AWTGraphicsConfigurationFactory'
|
|
|
|
|
|
|
|
|
|
|
| |
AbstractGraphicsConfiguration, if delegation is used.
This change restricts the usage of AbstractGraphicsConfiguration's getNativeGraphicsConfiguration()
to NativeSurface implementations and hence reduces complexity.
NativeSurface implementations are adapted and access to it's AbstractGraphicsConfiguration
is controlled via get/set method avoiding flawed usage (read/write), since read access shall
return the delegated AbstractGraphicsConfiguration, if used.
|
|
|
|
|
|
|
| |
instance pair
Don't use the shared device due to locking issues on X11.
Other platform impl. usually don't have native semantics on device/screen and may use it.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
public class GLDrawableFactoryImpl.
Michael Weber's test case (commit e7388512b69979d00e5a213a1b166a1b1726ae0c)
pointed me to the flaw in GLDrawableFactory which exposed a non-public method,
ie. 'getOrCreateSharedContext()'.
This lead to the assumption, that the 'shared' drawable/context of the factory is multi purpose
and may be used for application context sharing - which is not the case.
Changed Michael's test case to use a local shared pbuffer based drawable/context
and made the method non-public, where it belongs.
|
|
|
|
| |
creation on AWT thread (may deadlock)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
instance ; Applet's on OS X are working
Generalize OffscreenLayerSurface
- Using new OffscreenLayerSurface allows using this functionality in a clean manner,
ie. no 'dirty' usage of MacOSXJAWTWindow in a JOGL GL class.
- 'Promoting' OffscreenLayerSurface functionality to JAWTWindow
and it's handling to GLDrawableFactoryImpl::createGLDrawable().
- Move MacOSXCGLDrawableFactory's "MacOSXJAWTWindow getLayeredSurfaceHost(NativeSurface surface)"
to NativeWindowFactory "OffscreenLayerSurface getOffscreenLayerSurface(NativeSurface surface, boolean ifEnabled)"
Use local JAWT instance
- Only w/ a local JAWT instance per JAWTWindow it is possible to switch between
offscreen-layer and onscreen. We also have to determing offscreen-layer lazy
at surface lock, since only at that time we have knowledge whether it's an Applet
or not.
+++
ContextUpdater:
Use local pthread mutex, add DEBUG output
JAWTWindow/NewtCanvasAWT:
Adding methods to request offscreen-layer-surface (if supported),
besides 'if applet' this may trigger the new functionality.
+++
Applet's on OS X are working:
- OS X 10.6.4
- Safari:
- Hangs for a while at start .. whole screen freezes .. approx. 10s
- Sometimes crashes when Applet stops - after all our resources are released!
- Keyboard input isn't assigned sometimes.
- Otherwise .. works well, incl. offscreen/onscreen parenting
- Firefox 8.0:
- Hangs for a while at start .. whole screen freezes .. approx. 10s
- Sometimes crashes when Applet stops - after all our resources are released!
- Keyboard input is never assigned.
- Otherwise .. works well, incl. offscreen/onscreen parenting
- OS X 10.7
- Safari:
- Sometimes crashes when Applet stops - after all our resources are released!
- Keyboard input isn't assigned sometimes.
- Otherwise .. works well, incl. offscreen/onscreen parenting
- Firefox 8.0:
- Sometimes crashes when Applet stops - after all our resources are released!
- Keyboard input is never assigned.
- Otherwise .. works well, incl. offscreen/onscreen parenting
|
|
|
|
|
| |
It turns our that the native ContextUpdater does not work reliable in all cases,
hence we need to verify if the drawable size has changed as well.
|
|
|
|
|
| |
MutableGraphicsConfiguration allows derivations to set the chosen capabilities,
hence it is a non public API from which eg. JOGL GraphicsConfiguration derive.
|
|
|
|
|
|
|
| |
- AnimatorImpl: evaluate double negation
if(!skipWaitForCompletion(Thread)) -> if(blockUntilDone(Thread))
- Simplify finishLifecycleAction() and reduce timeout
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
I was too fast introducing these accessors, since they are currently not needed
and hence redundant.
Complete
- 7bc4c218b47033cb66f4eb5e707a86a0a4e60cff
Partial in regards to the holder interfaces only:
- d8fa00d35a49f4faf5f04aeb7e2bba4e972965f5
- f51e3dad6c4bd1f6d0001cecf6a0f692400ed602
- 46542168d64b37f544f61802693f15b59b224e4e
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
attach/detach our GL one
- it seems to be more stable, having one root CALayer attached to the JAWT_SurfaceView forever
- tackles crach at GL layer destruction
- proper release of all GL layer resources
- now final [gl-layer dealloc] happens at very destruction of JAWT object,
even though it was removed from root-layer earlier (and all other references)
- see comment in MacOSXWindowSystemInterface-pbuffer.m :: createNSOpenGLLayer(..)
- at least no more crash .. and resource release ASAP, but the GL-layer itself (see above)
|
|
|
|
|
|
|
|
|
|
|
|
| |
While impl. recreation for offscreen surfaces @ WindowImpl.setSize(),
commit 51ad6992e068f25d86d2c9e085bd7ec6f49d2432, it appears Animator deadlocks in some cases.
finishLifecycleAction() [blocks until pause() .. is accomplished]
queries whether we are even able to wait using skipWaitForCompletion().
The latter method can only query if we are on the animation thread or AWT-EDT, etc,
but not if we have traversed through a 3rd party thread, eg NEWT EDT.
This patches limits the wait with a timeout of 20frames @ 60Hz (20*16ms).
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
conflicts with setSize()
This fixes the NEWT OffscreenWindow conflict w/ setSize()
which represents the action of changing the window's size.
SurfaceChangeable's surfaceSizeChanged() merily notifies the actual size.
JAWTWindow: Remove setSize() since it propagates the surface size upstream only.
MacOSXJAWTWindow is not SurfaceChangeable complete (no surfaceSizeChanged).
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Better JAWT* name for offscreen layer surface:
- MacOSXCGLDrawableFactory, JAWTUtil, JAWTWindow, MacOSXJAWTWindow
- FIXME: Need to get rid of the cached JAWT instance,
in case we like to have dual usage of offscreenLayerSurface and onscreen.
This would be done implicit by using NEWT .. hence low prio.
AWTGraphicsConfiguration:
- Fix create(): Use capsRequested for AWT aligned caps if capsChosen is null.
(was capsChosen .. which is null)
- Add updateGraphicsConfiguration() which allows to update the AWTGraphicsConfiguration.
NewtFactoryAWT:
- Add updateGraphicsConfiguration() .. entrypoint for AWTGraphicsConfiguration.updateGraphicsConfiguration()
NSOpenGLLayer Impl:
- For 'some reason' the layer's position is initially negative, fix it @ 1st 'draw'
- Re-add CVDisplayLink OpenGL setting .. for what it's worth .. I don't know
JAWTWindow:
- Remove test setting: Only enable offscreenLayerSurface for applets if avail.
(New unit test enables it seperatly)
NewtCanvasAWT:
- If NEWT child is offscreen, attach AWTMouseAdapter and AWTKeyAdapter
to route these AWT input events to NEWT.
- Don't loose-focus if NEWT child is offscreen.
- Impl. NativeSurfaceHolder, NativeWindowHolder
- NativeWindow is created at construction
and it's GraphicsConfiguration updated at addNotify(..).
- At addNotify/reparent: try harder to determine proper NEWT child size:
- use preferred size if set
- use minimum size if set
- subtract insets from container size
OffscreenWindow/WindowImpl:
- Allow setSize() .. currently NOP for offscreen.
- WindowImpl: Commented out recreate case for offscreen-setSize ..
TEST: com.jogamp.opengl.test.junit.newt.parenting.TestParentingOffscreenLayer01AWT
Passed tests:
- GLCanvas
- NewtCanvasAWT/OffscreenWindow
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- NativeSurfaceHolder interface allows access to it's implementation's NativeSurface reference.
- NativeWindowHolder interface allows access to it's implementation's NativeWindow reference.
Implement or extend interface w/ these accessors whereever it seems fit:
- NEWT Window
- GLDrawable
These accessors allow lower layers to retrieve the NativeSurface/Window instance
w/o knowledge of the 'upper' frameworks, ie. NativeWindow/JOGL or JOGL/NEWT.
Commited in followups (due to mixed semantical changes in those files):
- JAWTWindow
- NewtCanvasAWT
- Some NEWT impl.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Undo 368cbf4462d7f3635c1ef4497424c360b5ccc203:
- "use SurfaceUpdateListener() to notify layer", simply use MacOSXCGLContext's swap impl.
MacOSXCGLContext:
- NSOpenGLImpl.create():
- issues createNSOpenGLLayer() using tex-size (maybe POT) as args
- attaches the NSOpenGLLayer to the JAWT layered surface host
- NSOpenGLImpl.setSwapInterval() propagates interval to our NSOpenGLLayer impl.
- NSOpenGLImpl.swapBuffer() (layer case):
- waits for v-sync if enabled (v-sync), or until native 'draw' is finished (tearing)
using our pthread/ NSOpenGLLayer synchronization.
The latter uses CADisplayLink or CVDisplayLinkRef for v-sync synchronization.
- flushes our local ctx
- triggers our NSOpenGLLayer to 'draw'
- MacOSXPbufferCGLContext create() sets the texture size independently of pbuffer size (POT/NPOT)
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
features, used by pbuffer drawable before a current context.
- Also extract getNextPowerOf2() -> GLBuffers (remove redundancy)
|
| |
|
|
|
|
|
|
|
|
|
| |
- attributes +NSOpenGLPFANoRecovery +NSOpenGLPFAAccelerated
- use SurfaceUpdateListener() to notify layer
- swapBufferImpl() adapt to 0038e2d41825c22bdd18a7b86a8229a3fab674a3
- pbuffer: don't enforce POT tex-size to surface size, but pbuffer only
- FIXME: Check POT pbuffer/tex-size for X11/Win32/.. !
-
|