| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
lifecycle deterministic; Remove warnings
Fix shared ctx release [onMainThread]
- Releasing the shared contexts caused a freeze of about 10s from one of the shared release operations.
[NSOpenGLContext release]
- Thorough triage concluded the workaround to release the shared ctx on the main thread.
- Using enhanced GLContextShareSet, see below
Make GLContextShareSet lifecycle deterministic
- Programmatically control the lifecycle of tracked shared ctx allows us using 'hard' references.
- Features queries for isShared() and ofc unregister a share set if all are destroyed.
Remove warnings
- MacOSXWindowSystemInterface.m used 'long', where 'GLint' was requested.
|
|
|
|
| |
testing (def.: enabled)
|
|
|
|
|
|
|
|
|
|
|
|
| |
(client-space/top-level, child/parent)
At resizing a perent window w/ a NEWT OSX child,
the window position needs to be updated since it's absolute.
Re-adding sending *Changed notifications via the appropriate WindowImpl methods.
Turns out they are missing in some parent/child situations (fullscreen for example).
Native getLocationOnScreen0(..) queries totalHeight by it's own to have correct values.
|
| |
|
|
|
|
|
|
| |
We not only shall skip windowing system triggered repaint if another animation thread
is running, but also if the current thread is the animator thread.
This keeps the animator intervals stable while resizing.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
which only holds the 'update' state now.
Avoid calling updater() for every makeCurrent(), but if view's frame has changed only.
This solves the pixel flickering experienced on OSX.
- GLContextImpl:update() -> drawableUpdatedNotify() w/ comments
- ContextUpdater holds context, view, old view frame and the update state.
It doesn't issue NSOpenGLContext update() by itself, but allows querying and clearing the update flag.
- MacOSXOnscreenCGLContext impl drawableUpdatedNotify()
- register via ContextUpdater, and use it if available.
|
| |
|
| |
|
|
|
|
| |
positions to 1st window
|
|
|
|
|
|
|
|
|
|
|
|
| |
pos/size notifications
Newt/MacWindow
- remove redundant manual window-move/set-size code
- Use local getLocationOnScreen(..), fixes positionChanged(..)
- setFrameTopLeftPoint(..) use totalHeight (w/ insets)
- create: don't 'retain' the window reference (ref counter)
- close: release view,
- cache insets - to be used @ create
|
|
|
|
| |
MainThread Thread name[s]
|
| |
|
| |
|
| |
|
|
|
|
| |
transparency
|
|
|
|
| |
Memory Access) @ glDrawArrays()
|
|
|
|
| |
bottom-left to our top-left origin.
|
|
|
|
|
|
| |
demands; Add missing frame.validate() after change.
This is esential for OSX, but may also cause an undefined state on other platforms.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Regardless whether the window is a top-level or child window,
we set it's position w/ absolut left-bottom coordinated.
To do so even for an AWT parent component, we retrieve the screen position
by traversing through the AWT tree and adding up each parent's rel. position,
since the native view only reflects the AWT frame.
Note: OSX does not use native views for each AWT component.
- In case we reparent child -> top, we cannot orderOut() the ex parent,
but need to just call orderBack(..), otherwise the whole ex-parent frame
gets hidden.
- In case we close a child window (and reparent child -> top),
we need to remove the parent/child relation and orderOut(..)
before close(..), otherwise the window artifact is left behind.
|
|
|
|
|
|
|
|
|
|
| |
InputEvent (Mouse) ; Test confined navigation w/ GearsES2
InputEvent adds new MASK values: CONFINED_MASK and INVISIBLE_MASK, set at event creation allowing
convenient testing of these mods.
GearsES2 demonstrates the confined navigation testing the CONFINED_MASK
and if having his mode acting on mouseMoved(..) and reset the mouse position.
|
|
|
|
| |
cleanup in X11
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
visibility:
- set pointer visible or invisible
confined:
- confine pointer to window, or not
warp:
- set mouse position within the window
Implemented for X11, tested manually with TestGearsES2NEWT (see code for action keys).
TODO: Windows, MaxOSX and Android (limited)
|
|
|
|
| |
== default && req != given
|
|
|
|
| |
is respected in EGL
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
release for a specific thread
- marked deprecated
- used to perf measure situation on omap3, ie
figuring out where the low perf. in GearsES2 comes from,
our core JOGL code or the GL usage.
Turns out it's the GL usage itself, ie the shader - good.
- calls are commented out in the demos
|
|\ |
|
| |\ |
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | | |
work)
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
subarrays
- Add GLArrayHandlerFlat gives better distinction of semantics
- update sub-array VBO name, if parent's interleaved array initializes it.
|
| | | |
|
| | | |
|
| |/
|/|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
cycles 45% -> 5% (from GearsES2 100%)
- NIO direct access from Java is expensive
- default is now array-backed non-direct NIO,
which guarantees array useage for Java computation (especially the inverse calculation)
- only update Mvi and Mvit if requested in the first place
- moved all local matrices to float[]
|
| |
| |
| |
| | |
TODO: custom position and change position ?!
|
| |
| |
| |
| |
| |
| |
| | |
visible=true, even if size==0
We need to earmark visible=true in case the native window could not be created,
even if the size is currently 0.
|
| | |
|
| |
| |
| |
| |
| |
| | |
SwapInterval
- turns out that on android, we cannot run w/o v-sync .. (?)
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| | |
available, ignore the given ID
Fixes Omap3 EGLConfig's which does not have a native visual ID
|
|/ |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Graph Shader Simplification
- remove enable factor and 2nd 'discard' branch
- use build-in 'max'/'clamp' functions, supposed to be faster
Graph Shader 'a'/'b' redefined
- 'a' is 1-pass shader only
- 'b' is 2-pass incl. (1st pass + 2nd pass)
- Works well on ARM Mali-400 MP (Galaxy S2).
- Doesn't work on NV tegra2 (P1202: Texture's gl states do not match with shader's),
however 2-pass on mobile seems to be overkill for now.
We may create a workaround (switch shader ..).
GraphUI 2-pass demo;
- Propagate renderModes and texSize to UIShape's render(..)
- TODO: Remove GL dependency in UIShape, maybe use a callback or visitor model
- Adding GarpUI 2-pass launcher (Android and Standalone)
|
| |
|
| |
|
| |
|