PPSSPP Git (2013/04/08)

Via http://www.emucr.com/

PPSSPP Git (2013/04/08) is compiled. PPSSPP is a fast and portable PSP emulator for Android, Windows, Mac, and Linux, written in C++.

PPSSPP Git Changelog:
* Merge pull request #1221 from raven02/patch-2
Fragment Shader : use mix() for decal/blend
* Fragment Shader : use mix() for decal/blend
* Buildfix for iOS. Fixes #1220.
* Merge pull request #1217 from unknownbrackets/fbo-flicker
Adjust FBO decimation to optimize for creation
* Change FBO age back to 5, gets in a create loop...
* Adjust FBO decimation to optimize for creation.
But, keep aggressively decimating for a few frames.
* Merge pull request #1219 from unknownbrackets/mpeg-fix
Initial sceMpegGetUserdataAu() from JPCSP
* Initial sceMpegGetUserdataAu() from JPCSP.
Because I got the stub wrong, so now it does something.
* Merge pull request #1214 from unknownbrackets/pad-fix
Treat the keyboard and pads differently
* Treat the keyboard and pads differently.
We basically never want two pads active at once, but keyboard is
different. So, they need to be treated differently.
Fixes #1212.
* Merge pull request #1213 from unknownbrackets/dlist-cycles
Try to match GPU cycles to the CPU
* Use regular events for GPU for now, quicker.
There's too much latency in threadsafe events, causing tests to fail.
Might break games too. I guess they need to execute more often...
* Don't allow two drawsyncs to be in play at once.
* Slow down the GPU clock estimate.
This makes Fat Princess for example much faster.
* Make sure lists aren't reused before they complete.
* Delay GPU signals and waits to simulate cycles.
* Merge pull request #1215 from raven02/patch-1
Add "Use Media Engine" to pause screen as well
* Keep Fast Memory as unstable
* Add "Use Media Engine" to pause screen as well
* Merge pull request #1216 from unknownbrackets/alloc-func
Initial stab at implementing some new memory funcs
* Initial stab at implementing some new memory funcs.
Borrowed from JPCSP, kept simple. Need to do tests.
* Don't "DirtyShader" at the end of frame, instead just "stop". Fixes issues caused by vertex arrays left enabled still hanging around.
* Bigger stack buffer for gpu stats (yes need to do something better here).
* Don't use PPGe for debug overlays. Fixes fps counter in games where it didn't work before.
* Merge pull request #1209 from unknownbrackets/threadqueue-perf
Don't save/restore VFPU regs for threads that don't use them
* Don't save vregs for threads that don't use them.
* Optimize resched with a running thread.
Esp. if it would go back to itself.
* fragment shader: RGBA DECAL does make more sense this way
* Merge pull request #1200 from unknownbrackets/threadqueue-perf
Improve threadqueue performance
* Fix crash if queue is empty during savestate.
On load it would then crash.
* Dynamically size the thread queue, explicitly alloc.
* Replace the thread ready queue with non-STL.
It seems much faster.
* Centralize the thread ready queue logic.
This makes it slower for now, but should make it easier to optimize correctly.
* Merge pull request #1203 from raven02/patch-10
Show upto 8 recent item for Android/iOS
* Show upto 8 recent item for Android/iOS


Windows version --> http://www.sendspace.com/file/qk2hf8
Android Version -> http://www.sendspace.com/file/eb3lg9